Nvidia has proven the space is incredibly lucrative and Apple is best equipped for high end chip designs. Remember 10 years ago it was unthinkable for an ARM chip to compete with x86.
First Apple has to prove they have competitive designs. Apple Silicon GPUs simply do not compete with the efficiency of Nvidia's GPU compute architecture: https://browser.geekbench.com/opencl-benchmarks
Apple's obsessive focus with raster efficiency really shot their GPU designs in the foot. It will be interesting to see if they adopt Nvidia-style designs or spend more time trying to force NPU hardware to work.
I think performance per watt is way in Apple's favor, but raw performance is not.
That said, an M4 Ultra (extrapolating from Max and Pro) would likely compete with my 3090, and with 192GB of memory (for 10x the amount it should cost) will out perform my 3x3090 AI server. And honestly, cost less than my 3 3090s + rest of the computer + electricity.
It won't outperform a bunch of A/H 100s (or even a single one, or any other cards in the enterprise realm) though, but it will cost an order of magnitude less than a single card.
Careful when comparing performance and efficiency. As a rough factor power increases quadratically as you increase clocks on a design, so you can quite easily make a high performance design low power by under-clocking it. The same is not true for the reverse.
Sorry, I was coming at this from the consumer side (since apple is a consumer product company). The majority of LLM use (by consumers) is in inference, not training, so I'd hazard to guess, the majority of people would rather have inference machines than training machine.
That's not necessarily news, unless I am missing something. Craig made an indirect mention of this during last year's WWDC regarding the private cloud compute.
I detached this comment from https://news.ycombinator.com/item?id=43158188 so we could pin the latter to the top. What you posted was fine! I just don't want to take up extra real estate at the top of the thread.
Going to be watching closely - but cynically, a promise of investment (for avoidance of tariffs) only needs to last one news cycle until tariffs are no longer top of mind. Then it can be walked back without tariffs being imposed.
Maybe instead of saying the t-word tariff, US gov can charge Apple a special fee on each iPhone. They can call it something catchy, like say, a Core Technology Fee.
This should be the top comment. Apple are doing business the way business is done, just like last time. Results don’t matter, it’s economic policy via press release. Form over substance.
Results matter, it's not hard to imagine that Apple considers the real risk of its promise and market position of being the privacy option being undermined by their supply chain risks, and leverage being used against them by privacy unfriendly actors.
Results don't matter as much as PR, this is time when this is unfortunately valid. Just look at US elections.
Measurable results affect rational aspects of our minds, PR attempts to attacks directly emotions bypassing the former, ie to induce impulsive shopping.
Also, what actual security? Apple is as vulnerable as cheap chinese phones against state actors using 0days. Apple devices are still being stolen for spare parts, Apple doesnt secure each component AFAIK and thieves know this (very recent case with friend of a friend, they even knew how to bypass that built in airtag tracking). I haven't seen anything but very well crafted PR statements on this topic. All money-accessing apps on absolutely any phone are a security risk.
What's the added risk here? It's fine to "risk" almost the entire iPhone itself to be manufactured in China but the servers for some random AI features need to be pure?
Sounds more like technical marketing and the company will treat any decisions around it as a marketing exercise.
Apple's commented previously on why they build in China, and it's beyond just the pricing - the supply chain for every single part they use is in China and mostly in the same geographic region, so there's a level of flexibility there they couldn't get in the US. It wouldn't surprise me if it was genuinely a goal for Apple to manufacture more in the US - they're a notoriously privacy-focused (corporate, not end-user) company, and China's known for IP wandering its way off campus. They're not going to sacrifice the iPhone economics until the US option is actually viable, but I'm not surprised they keep kicking the tires on US manufacturing.
> "the supply chain for every single part they use is in China"
Not entirely true. Some of the highest value components in an iPhone, including the CPU/SoC, baseband, and the majority of OLED displays, are sourced from countries that are not mainland China.
It's mostly about cost and market access to China.
Most smartphone supply-chain for Samsung and Apple exist outside China -- primarily in Japan (camera, sensors), South Korea (DRAM/NAND, OLED), and the US (various ICs fabbed at TSMC in Taiwan). There are quite a few reliable estimates/teardowns showing that these three countries account for close to about 90% of iPhone BOM (bill of materials). That's one reason why Samsung's smartphone unit was able to pull out of China without much disruption back in 2019 -- ie, low dependence on China.
I feel that Apple has pushed this misleading narrative a bit too long to defend their massive China outsourcing.
> They're not going to sacrifice the iPhone economics until the US option is actually viable, but I'm not surprised they keep kicking the tires on US manufacturing.
Apple could, with its immense cash hoard and cash flow, _make_ the US viable, but it chooses not to because it'd rather take the easy way out and have China or India or $COUNTRY fund it and return money to shareholders. They've returned money to shareholders rather than invest it in US operations, by design.
This is a classic feint to protect Tim Cook's entire raison detre. He built his career on super high efficiency operations by outsourcing to cheap labor countries. It relies on the low-to-no tariff access to US consumer money.
And I don't care that it's better for their stock price; that's Apple's problem not mine as a US citizen. And even as an Apple investor I would rather the money be spent on US on-shore operations.
There is no way Apple as a public company could just burn cash getting everything made in the US, shareholders would revolt long before the money ran out.
> And I don't care that it's better for their stock price; that's Apple's problem not mine as a US citizen.
That is the shareholder’s problems. People like to think that their investments won’t go batsh*t insane overnight.
> And even as an Apple investor I would rather the money be spent on US on-shore operations.
Apple doesn’t even make all, or even most of its money in the states. Not all of its shareholders are American, if they went this route they could lose half of their revenue overnight (as other countries note the protectionism and tariff or simply forbid Apple products from being sold).
This is correct if you only care about Apple stock price. But consider that there are people who simply do not care about Apple's stock price. Like me, an Apple investor. I care more about the US's industrial base than I do about whether it goes up 5% or down 5% (or whatever, it's besides the point)
>Apple doesn’t even make all, or even most of its money in the states. Not all of its shareholders are American, if they went this route they could lose half of their revenue overnight (as other countries note the protectionism and tariff or simply forbid Apple products from being sold).
So? I don't care about their revenue, I care about the future of American industry. Having a bunch of cash hoarded by old people is irrelevant if it isn't reinvested in something I care about. And I don't care about your supercar or Nobu reservation, or if some fund returns an extra 2%. This is despite being a direct beneficiary.
Live by shareholder return, die by shareholder return; the US is not and shouldn't ever be geared to shareholder return over everything else. Apple and other companies have freeloaded off the US for far too long.
Again, Apple ceases to be much of a company at all if they go your route of being isolationist. It isn't even about the stock price: their revenue tanks, their ability to produce tanks, everything about the company is basically just decimated. You might as well say "I don't care if Apple exists or not". So blow Apple away and how does that help the future of American industry?
Juche doesn't work in North Korea, it isn't going to work in the USA.
They've actually been diversifying iPhone manufacturing away from China for a few years already. As of April 2024, 14% of all iPhones were already manufactured in India. That's around 30 million phones per year. And Apple plans to double their India manufacturing again by 2028.
> This should be the top comment. Apple are doing business the way business is done, just like last time. Results don’t matter, it’s economic policy via press release. Form over substance.
If the Trump administration has any competence, they will rub those old promises in Apple's face until Cook actually does something meaningful.
The whole Trump administration is all about form over substance, though. I would not expect Trump to do anything actually productive about it, as long as Tim Cook sings his praise (and pays his dues).
Maybe the next administration should keep up the tariffs (as Biden did to a degree). Cheap trade with China distorts the tech sector too. Jobs and Wozniak were the products of a system in which americans had to build products at home. Tim Cook is the product of a system where you can become a trillion company by hyper-optimizing foreign supply chains. Which is better?
Apple went bankrupt under Jobs and Wozniak and was saved by hyper optimizing foreign supply chain company Microsoft only to rise 10 years later by focusing on hyper optimizing foreign supply.
There was a lot more than that going on and I think you've pretty generally mischaracterized the main problem with the mid-80's era Apple—which had nothing to do with domestic manufacturing and everything to do with not delivering new products that people wanted, at a reasonable price. You can claim overseas manufacturing solved the pricing component of that, but that's not at all clear: other companies were manufacturing in the US at the time and still out-competing Apple.
I don't know about Microsoft, but I'm very clear that the "miracle" operated by Apple was exactly to perfect foreign supply chain at a time when Intel/Dell/HP and others were still heavily focused on the US. The quality of Apple products was already there since the beginning, but they had no way to compete with the PC market until they figured out Asian supply chains.
That timeline isn’t even close to accurate. Apple was doing quite well in the 80s when Jobs and Wozniak were there. In the early/mid 90s was when they started going downhill (well after Jobs was gone), and by that time they had already outsourced a lot of their manufacturing (computers in Cork, Ireland and Singapore, and motherboards/components in places like Taiwan).
China's economic power is certainly not rooted in their isolationist social policies. They're just as bullish about foreign investment as the US was at the height of the free trade era.
Having access to cheap oversea steel allows Americans to focus on building companies with significantly higher value-add. Onshoring low-value industries is a massive human capital waste and an easy way to depress wages.
It allows a very small portion of Americans to build companies with significantly higher value-add.
It destroyed the futures of a larger number of Americans.
Then again, why do we make the distinction "American"? If you have people who became unfathomably wealthy by shipping off strategic industries to the lowest bidder regardless of geopolitical implications, does nationality matter anymore?
Do "normal Americans" pay taxes? From the numbers I've seen, ~1/3 - ~1/2 of tax filers receive more money from the government than they pay. To them, "refund season" is a cause for celebration rather than a stressful event.
Withholding has everything to do with it. Why do you think $10 an hour comes down to 1200 instead of 1600/month?
You can choose to withhold more or less, but the default taxation on w-2's do generally give a bit of a refund. Better to take out too much when you don't need it than slam down a gigantic bill when at once a year later.
The on-average crossover between negative and positive net total federal income (individuals will differ because of individual circumstance beyond just income level) tax when taking into account refundable credits (most notably, but not exclusively, EITC) is a bit below the median personal income but not that far below it, so certainly lots of individual "normal" (by most reasonable definitions) Americans do not pay net federal income tax .
But even if they don't pay net federal income taxes, they probably still pay a net positive amount in a variety of state taxes, federal payroll taxes, and federal consumption taxes (e.g., gas tax.)
"Refund season" is mostly a thing because the default w2 withholdings are set at a level where you slightly overpay on each paycheque, to avoid a surprise tax bill at the end of the year.
The problem with taxes is that it's a prisoner's dilemma. You need global cooperation at some base level of taxes, otherwise companies move to more favorable tax jurisdictions in the long term and offshore from there, which would hurt the US even more. It doesn't have to be all-or-nothing, but any marginal dollar of increased taxes in one place will have some non-zero effect of encouraging the next investment dollar to be spent elsewhere.
To be clear, I do think capital gains taxes are criminally low in the US relative to income tax, so I'm not arguing in _favor_ of lower taxes. I'm just saying why raising taxes isn't a panacea.
Creating an underclass that relies on economic elites paying taxes rather than being economically independent because you want to optimize for "high value add industries" is a terrible long term strategy.
> tax those people appropriately and pump that money back into the economy
So make the US to be like a far less successful country? Kill your economy by increasing taxes? The US economy is singularly successful because it has incentives to build businesses - see YC.
Have you tried living in a country that doesn't encourage businesses? They are often great tourist destinations. I'm in New Zealand and too many ambitious young people leave here: we have an emigration problem because our economy sucks. The government fixes the economy with 30% immigrants (disclaimer: I love immigrants). I have many friends that are never coming back here except for holidays. I hate the New Zealand government incentives for businesses (taxation and regulation) and I can see no way to fix them. Even our "business" political party ACT is completely fucked (latest story - they will be selling everything profitable to overseas "investors" - destroying the economy).
Taxation incentives matter to businesses. Be careful what you ask for because the majority have little understanding and vote for the wrong incentives.
Even business owners don't seem to understand incentive systems that well. Perhaps game designers do?
However I believe that incentives need to be marginal. If you already have a lot perhaps you need a big carrot as your incentive? I don't know any billionaires that I can ask how they feel about taxation incentives: I reckon you are making assumptions about what you think they should feel.
What makes Tim Cook make the US more money?
Taxation cliffs are shit. In New Zealand our Green party decided that 1 million was enough. Why would you bother growing a business after you reached 1 million? Retirement? A business is defined as being about making money (albeit some people do run "businesses" for other outcomes - why is Warren Buffett still working?).
High marginal taxation is also shit IMHO.
The hard part is to design the incentives so that productive people build your economy for the benefit of everybody.
If a government discourages business then the economy is crap and everybody suffers. See other economies.
Few people understand the incentives of others, and few people understand how wealth is created for all: the hoi polloi dismiss the wealthy as vampiric money grubbers. Anyone who uses the word capitalist in a derogatory way has been brainwashed. Most everything that makes our economies work is invisible non-monetary rewards. https://news.ycombinator.com/item?id=43162596
I can speak for my own financial incentives. My perception is that I have an effective tax rate of well over 50% in New Zealand (any retirement savings are not safe because our demographics and governments will screw our economy).
I do not feel the incentive to work in a business - My attitude means I now produce marginally less than I could for the New Zealand economy (I still pay taxes so they are advantaged but they could get a lot lot more from me). I now mostly selfishly concentrate on those closest to me. Why should I work if it isn't marginally beneficial enough for me? I'm no more selfish than my retired friends that I know (a wide variety of people from many walks of life).
(Reëdited to expand and clarify).
We can't decide how much is fair. Compare yourself to a dead king - what is fair? We can design systemic incentives so that we each make the world better for everyone. Not that that it is easy... Trite thoughtless dismissals of the most productive members of society are not helpful.
Edit 2: I guess this discussion is as close to work as it gets for me. Too much adulting. Should I get into politics? Are morals an impediment to helping others? There are too few politicians I admire, and too many I wouldn't want to shake hands with or be associated with. Every idiot has political opinions - how much of an idiot am I? Every politician is smart enough to win an election - they are not stupid yet they make too many horrific mistakes. What about the cryptically smart ones? I see how systems affect people that join a system. What would I become if I join our political system? Understanding our different systems is hard because they grow so weirdly with vestigial complexities due to history, complex interactions, and reflexivity.
>Do Medicare and Medicaid exist without businesses?
In a purely technical sense, yes. Because you don't necessarily need an American salary to pay taxes that cover these facilities.
It was very much a concerted effort for most other non-govt Healthcare to be tied to often American jobs. Which of course causes a cacophony of problems when less employers are even offering full time work.
>Why do you look at money as though that is all that matters?
It does not, but business these days sucked up enough money that it's starting to affect basic survival, let alone any pursuit of happiness. There's no point finding upsides when the common person is is so low on the totem pole.
Making hypotheticals of "well look on the bright side, you're not dead" doesn't help either. When America starts using that wealth to make sure no one in a first person country isn't dying on the street, we can discuss the subtleties of capitalism.
>Every poor person I've met avoids taxes.
Well I can't speak for New Zealand. You can't tax a poor person with no income. That's how bad the situation is here.
> Because you don't necessarily need an American salary to pay taxes that cover these facilities.
That is a weirdly employee centric view. I'm talking about the US economy. American salaries depend on American businesses. America has some of the best healthcare available in the world. If US businesses are fucked due to the beliefs of citizens (or whatever else), then the US socialised healthcare is fucked too. There's plenty of poorly run countries to compare against (including Cuba where I discovered their lies about their healthcare first-hand as a tourist). NZ socialised healthcare is okay but our economy is not improving and regardless of our desires for more, the social benefits have no choice but to match our economic output.
> it's starting to affect basic survival, let alone any pursuit of happiness
Only if you're one-eyed. US citizens are the rich. In a fair world we would tax all Americans at 90% and redistribute that to the poor in the rest of the world. Maybe same for NZ too (Wikipedia shows that NZ's disposable median income is ⅔ that of the US however it also strangly says that NZ's median wealth is nearly double that of the US -- I'm guessing because houses are more unaffordable in NZ). Income is usually a better measure within an economy of useful output (economies can't really save for next year). The US federal poverty line is about $16000 for one person - a hell of a lot of money for people in many countries.
> Making hypotheticals of "well look on the bright side, you're not dead" doesn't help either.
I guess you're referring to my comment "Compare yourself to a dead king - what is fair?".
My obfuscated point is that few people (maybe narcissists) would give up their modern life to live in past poverty. Antibiotics, freedom, technology, access to the intellectual output of the world. We are mostly a lot better off than the past. Most people don't value that instead they are money-centric (as many of your comments are). Most people seem to compare themselves to people that are wealthier than themselves and then complain about how they are not getting their fair share. Few people compare themselves against the global poor and then talk about how much they should share their wealth downwards. They talk about how others should share their wealth - they rarely seem to consider how they should share their own wealth. Especially ironic given that it appears that the majority of commenters on HN are the wealthy of the world (and often part of the tech overlords - e.g. YC).
The US is often a parasite upon other countries. If you were to say that the US pays it back to poor countries with technology (mostly from rich companies), then you would be implicitly arguing that wealthy US companies deserve to be wealthy. I recall that weapons are the biggest US export (nice!)
I guess I'm saying is really take care not to kill your geese laying golden eggs (even if you think the geese seem to be keeping too much golden egg to themselves): the socialised good that you have depends on those geese (US businesses). The bad is bad but don't destroy the good.
An economy is a delicate balance - as shown by many failed economies.
> When America starts using that wealth to make sure no one in a first person country isn't dying on the street
So What Does Drive Political Disagreement?
If you’ve read The Psychopolitics Of Trauma, you already know my answer to this: it’s all psychological. People support political positions which make them feel good. On a primary level, this means:
\* Successful people want to hear that they deserve their success.
\* Unsuccessful people want to hear that successful people don’t deserve their success, lied / cheated / nepotismed their way to the top, and are no better than they are.
\* People want to knock down anyone who makes a status claim to be better than them.
People want to feel like their own identity group is heroic net contributors, and that their outgroup are villainous moochers.
People want to feel like their own identity group deserves more power.
* People want to feel like their preferred lifestyle and policies have no negative implications at all and they don’t have to feel guilty about them.
* People want to feel like they’re part of a group of special people poised to change the world, and everyone else is hidebound bigots who resist temporarily but will eventually be forced to recognize their genius.
People want to virtue-signal: demonstrate that they have the good qualities that their ingroup considers most important.
* But people also want to vice-signal: demonstrate their willingness to breezily dismiss the supposedly good qualities that the outgroup considers important.
No, the analysis (and it’s not exactly rocket science) says just the opposite: Way more downstream manufacturing jobs that rely on steel as input are lost, vs. domestic steel production jobs gained.
I get the value add argument, but lots of people just need income to pay for living expenses. Without an income, those people become disaffected and sometimes violent. Then they embrace right-wing protectionism because, while their gadgets are cheaper, they have no income to buy cheap gadgets.
Nor can they move to these offshore places (where the cost of living is lower) because immigration laws exist in part to control worker mobility.
> Having access to cheap oversea steel allows Americans to focus on building companies with significantly higher value-add. Onshoring low-value industries is a massive human capital waste and an easy way to depress wages.
That's the talking point, but it's bullshit. A lot of those "low-value industries" are fundamental capabilities, and China sure as hell isn't going to let the US own the "higher value-add" areas. They dominate those next, and the US free-trade business elites will be fine with it as long as they get to make some money for themselves.
Being a high value-add area is endogenous to how hard it is for others i.e. China to reproduce. In other words, if it were easy to make GPUs they wouldn't be so damn expensive.
> Being a high value-add area is endogenous to how hard it is for others i.e. China to reproduce. In other words, if it were easy to make GPUs they wouldn't be so damn expensive.
China's going to put the money into making GPUs, and they're going to get it right, probably sooner than later. Then they'll drive the American manufacturers out of business, like they've done in many areas before. Their government isn't beholden to the profit-focused capitalist attitude that is one of the West's biggest vulnerabilities.
Also, it's pretty foolish to 1) forget power and security doesn't come from rarefied high value-add stuff, 2) that there are a lot of people that can't be employed doing stuff like making GPUs.
>> Isn't this why we declared independence in the first place? To get away from the British restricting free trade?
No. I'm not sure where you got that idea. If you look at something like the Boston Tea Party, it wasn't high taxes on tea that were being protested against, it was lowered taxes on tea that undercut the smuggling operations of people like Sam Adams and John Hancock. "No taxation without representation" makes better press than "No undercutting my smuggling operation" though.
In the early years of the US, between 80 and 90 percent of federal revenue came from tariffs. Not exactly free trade.
> In the early years of the US, between 80 and 90 percent of federal revenue came from tariffs...
To be fair, the Federal Budget back then was 2%-ish of GDP. And their political consensus gave the Federal Gov't very few things that it had the power to tax.
Govt spending as a share of GDP is probably a good measure of how involved the government is in the economy. There are arguments that too much government involvement leads to stagnation, which considering many of the economies the US regularly out-grows have higher share of govt spending as a % of GDP has some merit. Its an interesting economic question what level that would be which changes depending on how you approach the problem.
You’re incorrect about history. Mercantilism not only restricted foreign trade, but restricted domestic industrial development by requiring the colonies to sell raw materials to Britain and buy finished goods from the Britain. Tariffs were a core pillar of the Lincoln Republican Party.
There’s been an isolationist wing in tech as long as I’ve been in it (early 2000s). I remember chatting with someone at Cisco/Juniper in the late aughts about Huawei ripping off their router designs down to the silk screening. Of course today Huawei makes their own state of the art routers with their own silicon, and some lower-end Cisco/Juniper gear is white boxed foreign equipment. And of course tech folks were complaining about immigration and outsourcing back in the early 2000s when Republicans were enthusiastically supporting both.
I want to say “that’s not what isolationism means”, but I realize it starts to feel vague just like the word “fascism”, used when convenient but varies wildly in rhetorical meaning… to be more specific is better, I like what George Washington had to say about it in his farewell address because it shows the nuance of the topic across the spectrum, it’s not as simple as isolation good vs bad:
The great rule of conduct for us in regard to foreign nations is, in extending our commercial relations to have with them as little political connection as possible. So far as we have already formed engagements let them be fulfilled with perfect good faith. Here let us stop.
Europe has a set of primary interests which to us have none or a very remote relation. Hence she must be engaged in frequent controversies, the causes of which are essentially foreign to our concerns. Hence, therefore, it must be unwise in us to implicate ourselves by artificial ties in the ordinary vicissitudes of her politics or the ordinary combinations and collisions of her friendships or enmities.
Our detached and distant situation invites and enables us to pursue a different course. If we remain one people, under an efficient government, the period is not far off when we may defy material injury from external annoyance; when we may take such an attitude as will cause the neutrality we may at any time resolve upon to be scrupulously respected; when belligerent nations, under the impossibility of making acquisitions upon us, will not lightly hazard the giving us provocation; when we may choose peace or war, as our interest, guided by justice, shall counsel.
Why forego the advantages of so peculiar a situation? Why quit our own to stand upon foreign ground? Why, by interweaving our destiny with that of any part of Europe, entangle our peace and prosperity in the toils of European ambition, rivalship, interest, humor, or caprice?
It is our true policy to steer clear of permanent alliances with any portion of the foreign world, so far, I mean, as we are now at liberty to do it; for let me not be understood as capable of patronizing infidelity to existing engagements. I hold the maxim no less applicable to public than to private affairs that honesty is always the best policy. I repeat, therefore, let those engagements be observed in their genuine sense. But in my opinion it is unnecessary and would be unwise to extend them.
>when belligerent nations, under the impossibility of making acquisitions upon us, will not lightly hazard the giving us provocation; when we may choose peace or war, as our interest, guided by justice, shall counsel.
Ironic that as a Canadian, the US is moving from the nation that would be guided by Justice into the belligerent nation in this situation.
It also serves as a lesson to us that we should have learned from you and George Washington, and stood on our own first and ensured our own security before cooperating with others. We have a long way to go to get back there now, unfortunately under the position of potentially our closest ally and economic partner being belligerent, untrustworthy and unreliable.
I believe the body of people who are promoting such a narrative (and it’s clearly coordinated) have major conflicts of interests, they want to maintain the status quo for their own enrichment at the expense of any specific nation, because in reality plain old paying attention shows nearly the exact opposite is true in every sense.
They are very good at their propaganda which is exactly how they got in that position but they are not looking out for your interests in the slightest, they just want to “manufacture consent” for the forward march of the global hegemony, that oligarchy… nobody in their right mind, with eyes unclouded by hate, would come to these conclusions naturally.
It’s probably the scariest thing too, and it’s nothing new! Go read Chomsky’s book by the same name “manufacturing consent” and he lays out many examples that were happening in the 70s-80s, and they are following the same playbook today just with Ukraine and Gaza instead of Colombian Jungles and Vietnam.
>I believe the body of people who are promoting such a narrative (and it’s clearly coordinated) have major conflicts of interests,
I'm a Canadian. Operating SOLELY on the actions and statements of your executive branch, not the media's reporting but the wording of the government, their executive orders, their direct public statements. your government is increasingly belligerent, untrustworthy and unreliable.
Belligerant -> Constantly making subtle threads of annexation. Calling the Prime Minister of Canada the "Governor" of Canada. Constantly lying about trade deficits that are surpluses and drug and migrant problems that are actually a bigger problem moving north across the border than south.
Untrustworthy -> After renegotiating NAFTA to USMCA and hailing that as a great agreement, now its a shit agreement and he's putting tarrifs on to get more from Canada under the threat (and likely actuality) of causing economic harm to both our countries.
Unreliable -> You were our biggest and staunchest ally. Electing an administration that is actively hostile to our government and sovereignty means you are no longer reliable as an ally.
> they want to maintain the status quo for their own enrichment at the expense of any specific nation, because in reality plain old paying attention shows nearly the exact opposite is true in every sense.
The current actions of the administration are destroying hundreds of thousands of lives in order to enrich a few thousand at the top. They are alienating ally's, destabilizing peace and almost guaranteeing more conflict and war in the near future.
Chomsky warned of the damage a demagogue supported by a disenfranchised, hurting and angry populace can have on a country. This is happening now, and half the country is burying their head in the sand claiming its propaganda.
>This is total nonsense, nobody thinks they can annex Canada…
Trump defies typical expectations and does things everyone else thinks would never happen. Often because to do so would be very damaging and idiotic, but that never actually stops Trump.
However you do NOT threaten to annex, take over or otherwise threaten the sovereignty of an ally or friend. That is geo-politics 101, so regardless of how serious he is about doing it, the act of threatening it is belligerent and shows he isn't a reliable trading partner or ally.
Outside of annexing, the trade war is also actively hostile on top of being perpetuated on complete lies.
>they just want to “manufacture consent” for the forward march of the global hegemony, that oligarchy… nobody in their right mind, with eyes unclouded by hate, would come to these conclusions naturally.
The oligarchy has taken over. The few that benefit from the "global hegemony" you refer to, which is largely the 'interests of the rich' are now completely in control of the US.
The US administration absolutely is not looking out for the interests of the average Americans. Most of what they are doing directly hurts most or all Americans in one way or another, and the few things they do that help Americans benefit the rich the most.
Now is the time much of the US populace has always claimed it would stand up against tyranny, an oppressive "lord" class and kings. They are watching it and cheering it on instead.
Apparently, yes. I saw mention of discussion around the Trump administration potentially giving Apple a tariff waiver. And I believe in Trump’s last term, Apple did have some sort of waiver.
I’m on mobile but Googling for “Apple tariff waiver” and “Apple tariff exemption” will point you to several news items.
There is no jobs problem in the US though. Unemployment is at 4% which is mostly just job churn. Long term unemployment is only 1%.
US consumers, that’s all of you, are being hammered with taxes on imported goods most of which can’t realistically be produced in the US anyway, to solve a problem you don’t have.
A commitment like this takes years to plan. It can’t possibly be a response to tariffs announced weeks ago. This is all optics.
The government determines the employment rate via surveys, i.e. they just go and ask people if they're employed. It's not a calculation from taxes or from employers or anything.
So it's up to the gig workers if they think they're employed or not. Presumably this depends on how often they do it.
False, USA has a big problem with manufacturing. All US jobs are service jobs to prop up consumer economy, that have no strategic benefit.
A lot of fake employment and low productivity jobs are in the government/NGO sector, paper pushers, DEI jobs, law/compliance type jobs - that should have been manufacturing jobs instead.
USA has no shipyards and infrastructure is crumbling precisely because of misallocation of resources and labor
How is the Jones Act responsible for the failure of domestic ship building? Seems like the Jones Act didn't go far enough if we really cared about a strong domestic ship building industry.
if shipbuilders were not insulated from competition, they would have offshored their manufacturing long time ago and sold out to Hyundai or some other foreign conglomerate. (just like the rest of manufacturing left USA in 80s-90s)
Jones Act is the reason US has at least some form of domestic shipbuilding
The US domestic civilian shipbuilding doesn't exist. It's dead and already smelling.
For example, the Washington state needs new ferries, and their cost is literally almost 10x than the cost for the similar ferries produced in Turkey for the nearby Vancouver, BC. With this kind of cost gap, there is simply nobody who would buy the US ships unless they _have_ to.
Without the Jones Act, the shipbuilders would have adapted long ago to produce high-margin high-tech components for the ships for the international market. And likely in larger quantities than they do currently.
I fully believe that the real ("main street not wall street") economy is in worse shape than government numbers on unemployment suggest and both sides are to blame for different aspects of this problem.
But nothing Trump is doing is going to fix your situation.
In fact he (or rather Elon/doge) is very actively making things worse for you with the massive government layoffs, flooding the market with even more people to compete with you for jobs making finding work more difficult and also eventually dropping all of our wages.
>But nothing Trump is doing is going to fix your situation.
I'm aware. I'm sure he's responsible for at least 3 job freezes I ran into mid-interview this year. He's literally costing me job opportunities because no one can budget around this chaotic government.
4% too many and probably understated. The BLS repeatedly underestimated unemployment during the previous administration. Also the labor participation rate, which is harder to game, still hasn't reached pre-Covid levels yet: https://www.bls.gov/charts/employment-situation/civilian-lab...
In February 2020 it was 63.3% and in January 2025 it was 62.6%, for a difference of 0.7%. Also note the steady decline post-2008 and the multi-year plateau that jitters around 63%.
Having the plateau change from ~63% to ~62.5% isn't an unreasonable scenario.
Yes this is one reason tariffs are so valuable to a corrupt POTUS. They have essentially unilateral and very fine-grained control over them, down to exempting specific companies or products outright.
Congress needs to step up on this, honestly. The entire idea that the President can unilaterally implement trade policy is as plain a violation of separation of powers I can think of, and SCOTUS is a fan of non-delegation doctrine.
iirc as soon as anything becomes beyond the border the President holds the keys for various reasons including the ever-vague “national security” but also due to being prescribed as the primary negotiator https://en.m.wikipedia.org/wiki/Treaty_Clause
Tariffs are not treaties, they're taxes/duties - only Congress has the power to raise them. Article I is extremely clear on that. Historically, tariffs were always raised by acts of Congress, not by Presidential fiat.
Trump is using extremely misguided legislation from the 1960s/70s where Congress allowed the President to enact tariffs for national security and emergencies. There is a very strong argument (in the sense it resonates with the conservative SCOTUS majority) that Congress cannot delegate its fundamental powers to the executive by legislation alone.
I think people are just too cowardly to bring a case in front of the courts to challenge the constitutionality of it all. Non-delegation doctrine is what the Federal Society want to use to kneecap all federal regulation. Trump operates on a spoils system so it's not in the interest of conservatives or businesses to challenge him, for fear of retribution.
Trump is using tariffs not to raise revenue, but rather use it as a stick to force companies to invest in USA.
Previously they were outsourcing and offshoring as much as they could get away with it. Which led to transfer of advanced technologies outside USA and America losing its manufacturing and technology edge
So how's that going? Outsourcing seems to be going strong, the tarriffs instead pissed off allies who are preparing counter-tarriffs, and the CHIPS Act is being dismantle as we speak (there goes our investment.
Tariffs are a form of taxation. If I want to import say tea, and the government is placing a tariff on that imported tea, I am effectively taxed by the government. And only Congress can impose new taxes.
Not saying you're wrong, but... I have seen claims that tariffs are a source of government income that Congress doesn't control. You're claiming they do.
I haven't seen a citation from either side. Can you substantiate your position?
I have already explained my thinking up this comment chain. I'm mostly replying to GP who misunderstands that the intent of the tariffs is besides the point.
TL;DR read Article I section 8, read up about the Trade Expansion act of 1962 and Trade act of 1974, and "non-delegation doctrine", and you can trivially find legal debate about the constitutionality of IEEPA. Rather than listen to random nerds on HN you should seek out this information yourself.
Clearly a non-sequitur, but I'll bite anyways. Paul Pelosi is a VC, and I'm sure that most of the Pelosi net worth is due to his income, not hers. Since the laws preventing members of Congress from trading on information they receive as part of their duties, you can't say that the Pelosi's have violated any securities laws.
And both sides of the aisle benefit from this. Whether it's legalized insider trading or jumping to corporate jobs when out of office, it's a corrupting influence. All members of Congress, SCOTUS and POTUS should have to place their assets into blind trusts. That won't stop this corrosive influence, but it is the bare minimum.
> Since the laws preventing members of Congress from trading on information they receive as part of their duties, you can't say that the Pelosi's have violated any securities laws
The feds can and do go after people for using family members and friends to execute trades. So if Nancy Pelosi told her husband some material non-public info, and Paul Pelosi traded on it, that would still be insider trading.
There was a guy at Microsoft who was caught once using a friend to place trades. He said he talked himself past his ethical concerns by reasoning that members of Congress do it.
Edit: to be clear, the absence of a prosecution does not mean that the Pelosis did not insider trade. Nor that they did. We can't tell from this distance, only speculate.
"Paul Pelosi, 83, sold 30,000 shares of Google (GOOGL) stock in December 2022, just one month before the tech giant was sued over alleged antitrust violations."
> Since the laws preventing members of Congress from trading on information they receive as part of their duties, you can't say that the Pelosi's have violated any securities laws.
Legislators step up when enough of their voting constituencies make it clear that they value something as a non-negotiable (assuming votes still matter).
Which means those who care about this are back to not only contacting legislators but also persuading a lot fellow voters that separation of powers is crucial and worth prioritizing over familiar well-handled and loved heuristics.
Yes a million times. For all the rhetoric about authoritarians, the Democrats never seem to want regin in Executive power when they are they majority. It is like a game of chicken where America winds up with a populist dictator from either the left or right.
Yes because populism is a reaction to government being generally unresponsive to people’s needs.
Congress has become increasingly unproductive and unresponsive. There are many popular policies that Congress essentially ignores, and many problems that go unsolved. So trust in government dwindles and people crave strongman solutions.
I’m not sure there is a solution. There are so many interlocking problems gumming up the process that any “we just need to fix X” solutions (where X is gerrymandering, money in elections, lobbying, the two party system, first past the post, corruption, income inequality, the electoral college, the slow death of journalism, consolidation of industries, etc) are nearly impossible and also probably insufficient because they all feed back into one another: they are both causes and effects.
So when people are mad about a downstream effect like the price of eggs and digging any deeper touches one of the topics above (“to fix egg price gouging you need to reinvent the political system” sounds a lot like “to make an omelette first you need to create the universe”), it’s really easy to throw your hands up.
A codebase accumulates messes during it's entire lifetime. Sometimes, it gets to the point to where a rewrite is a good option. Even if the rewrite doesn't have all of the features...Even if possible, it may be better not to keep all of the features. As some features aren't important enough to maintain. The same can be said about government structures.
That is a fair analogy of the problem. But the solution is inherently more complex.
1. The agents of political change are themselves subject to politics. (Unlike maintainers of a codebase, who are able to make top-down decisions about the code.) If you make a bad decision, it may be the last one you’re allowed to make.
2. There are no unit tests for politics. There is only history, which is an imperfect predictor.
Granted large political systems are uniquely complex. But complex & old software can share the attributes you outlined.
1. Politics does enter into programming teams. There's several articles of this genre. Where the most highly regarded programmer made bad decisions over a long period of time & exhibited hubris in doing so. This programmer was fired, but if he was a founder otherwise held some leverage, then he may not be fire-able. However, other staff can leave. It's more difficult to leave the impact of politics.
Entire industries can make bad decisions due to incentives. Such as I must learn technology X because companies hire for technology X. Companies choose technology X because developers know technology X. Technology X could have fundamentally flaws. Compounding complexity in the ecosystem as additional tech is made to fix the flaws...which will also have flaws.
These cycles of compounding bad strategic decisions yet good tactical decisions can last decades.
2. In some ways there are unit tests for politics. Geopolitical signals can be sent. So if X happens, there's a tacit agreement that Y will be the response. Unit tests are great for small units. However, complex systems can be difficult to comprehensively test. This is why there is fuzzing...which is effectively a monte carlo simulation. Monte Carlo simulations are widely used in political analysis.
3. There's non-deterministic inputs. And the system can be more complex than what's possible to model. Leading effectively to non-determinism in making decisions about how to modify the system.
Charitably, tariffs exist so POTUS can either lower taxes or increase jobs in US, but both would take time to pan out assuming things go well. So if a company is willing to onshore money or jobs, its achieving its intended purpose in their eyes.
No I think Mexico/Canada were largely about stopping immigrants and fentanyl smuggling. But China was targeted for not doing enough to stop manufacturing of fentanyl precursors.
Sorry to say but that's already been walked back on after Canada committed $1 billion dollars for extra northern border security and it made no difference in the tariffs discussions.
- According to CIS, the number of Canadian crime groups producing synthetic drugs doubled between 2023 and 2024
- There's a lack of Canadian agents who are tasked at preventing this and current legislations make it very inefficient between federal and provincial law agents
- There's an upward trend in Fentanyl seizures in Canada the last 2 years
- Fentanyl is now being produced domestically in Canada
All of that is within the control of Canada with better policies.
Let’s put it into perspective, because those numbers don’t give a baseline for what the problem is. Also they don’t necessarily have anything to do with trafficking.
Last year there were 45 lbs of fentanyl intercepted crossing into the US from Canada. Thats a backpack. There’s 500x as much coming from Mexico.
It’s unrealistic to expect that zero fentanyl will come into the US from Canada, and until that happens we will tariff all trade with them.
Zero fentanyl is a fantasy and will never be achieved. That thing is way too small and can be carried around too easily*
Tariffs are just a lever to get things done on the international front.
Canada has been neglecting security for a long time, so it's a wake-up call, and it's not a bad thing to put this out publicly to change things.
CBSA only inspects about 4% (some sources say even 1%) of all containers that ship in Vancouver. Quebec-Vermont* border control has been a joke for years.
Sure, Trump is using all kind of tricks to put pressure against his trade partners to secure some wins that will solidify him as a change agent to boost America, which will appeal to his electoral base. At the same time, it may actually bring good results to the US economy, bringing major investments in the country and negotiating better trade deals.
You may not agree with the means to get there, but you can't deny there's an argument to be made about his "America first" policies and why it could benefit the average people in the long run.
So therefore that allows the President to go back on a trade agreement he personally signed in his last term? I'm not going to disagree Canada should do more about reducing Fentanyl, nor that Canada can't control it with better policies. I am not clear on why this allows the United States to go back on agreements and allows the President to threaten with tariffs that seem to change weekly.
I am a free trader in principle. However you have a country (China) with an authoritarian government that makes favored industries subsidized.
Of course the standard economic argument is that China using its GDP to make goods cheaper for our own citizens to purchase is better for us - they are subsidizing our economy. However it ignores the strategic disadvantage by our country losing its manufacturing capabilities.
The graphs may show economic advantage. It’s hard to quantify the long term strategic and militaristic disadvantage to not being able to make anything yourself if a world war occurs.
"The company responded by intensively lobbying the U.S. government to intervene and mounting a misinformation campaign to portray the Guatemalan government as communist.[18] In 1954, the U.S. Central Intelligence Agency armed, funded, and trained a military force that deposed the democratically elected government of Guatemala and installed a pro-business military dictatorship.[19]"
> However you have a country (China) with an authoritarian government that makes favored industries subsidized.
This is overlooking the forest for one tree. The thing is, mean chinese manufacturing wages are $25k/year (purchasing parity adjusted! $15k unadjusted) for a 49h week.
That is the reason that so much manufacturing/industry has shifted there, not some nebulous "Chinese government subsidies" (not saying those are not a thing, just that they don't really matter all that much).
> It’s hard to quantify the long term strategic and militaristic disadvantage to not being able to make anything yourself if a world war occurs.
Certainly. But forcing low-skill industry to stay at a relevant size in a high-wage country is expensive business (compare agriculture, which is subsidized basically for exactly this reason) and not straightforward (see Jones act).
Presenting tariffs as a viable alternative to taxation is just beyond ridicule, but that has not stopped people so far either...
Absolutely. You do need a minimum baseline for infrastructure, government stability and workforce.
Most of Africa is just starting to slowly get there, Bangladesh is already very relevant for textile production.
I would expect the same basic trend to repeat that we saw with electronics manufacturing in 90s Japan:
First cheap products move (very wage sensitive), then the local sector expands, wages rise with the whole local industry moving up the value chain, then at some point local wages become high enough for the whole process to repeat with the next low-wage country...
I think trying to block this trend off with tariffs is a futile waste of taxpayer money which american consumers are gonna pay for.
Spending tax money to keep some degree of self-sufficiency in critical industries (like with agriculture) can be a solid idea if done sparingly and cleverly, but that is not how the current US admin has approached this...
Salaries are just a small part of the reason industry works in China.
The bigger picture is that China invests in the development of an industrial chain. This has many aspects: infrastructure, education, training, housing, and of course tax incentives. The USA decided to stop investing in practically all of these. Even scientific research, the last area in which the US used to lead, is now in jeopardy from both sides: competition from China and internal cuts.
I'll concede that having a solid baseline of infrastructure, (political) stability and a motivated/educated workforce is necessary, and China did well in building this up.
But I strongly disagree with your conclusion.
Lets assume that the US did all of those perfectly:
- Brilliantly educated workers with the perfect ratio of industry specific knowledge/experience
- Cheap housing near industry hubs built by the state
- The best and cheapest to use ports, roads and railway networks on the planet
- No tax on manufacturing workers income
Some of those are ludicrous/unrealistic for the US to provide.
But even if you managed to do all this-- that still does not make US manufacturing industry competitive. Because those US workers will still want a locally competitive wage instead of <10$/hour.
The reason the US is not competitive is exactly because it doesn't spend the money needed for that. China did it. To be more concrete, if the government spends money to build housing, workers don't need to pay so much to have a home. If the gov spends money on public transportation, workers don't need to buy expensive cars just to get to the job. If it spends money on free health care, then workers don't need to pay for expensive insurance. If the US spent money on (near) free higher education, workers wouldn't need to pay high costs on student loans. These are all items that make the US uncompetitive with other nations.
> The reason the US is not competitive is exactly because it doesn't spend the money needed for that.
You are making some kind of logical jump here that I can not follow. I just listed basically the absolute best that the US could have ever done-- but even in that absolute dream scenario (tax free income for manufactoring workers? I mean in what world do you see something like that ever actually happening?), the US is still not competitive in a direct comparison, because US workers have just no reason to work manufacturing for 10€/hour (when they make ~30/hour right now).
You can stack all the incentives you want-- the gap in wages/standards is so large that apart from straight up paying the difference (in either tariffs or subsidies, and that is a lot of money), you are not going to make US manufacturing competitive in a head-on comparison.
What you forget is that is exactly these inefficiencies that inflate US salaries. You need to make more money just to survive in most US cities, which forces companies to increase salaries.
> What you forget is that is exactly these inefficiencies that inflate US salaries. You need to make more money just to survive in most US cities, which forces companies to increase salaries.
I think you are reversing cause and effect here. Wages are not rising because things are expensive-- things instead become expensive because people are "rich" and can afford them.
I suspect (not an accusation!) that you would intrinsically like to see the US run healthcare and education in a government controlled way, at-cost, instead of allowing excessive private profits there (which I think is a good idea!).
But advocating for such changes in the name of making US manufacturing competitive is dishonest in my opinion, because I absolutely do not see those shrinking the wage-gulf sufficiently for US factories to compete head-on with China.
Furthermore, I don't even think you want to be competitive with China in this regard. Having a significant percentage of Americans working in/for factories to produce simple goods for 10$/hour strikes me as a step back, even if you would bundle this with a bunch of positive progressive improvements.
Sour grapes. Most economists were just happy with this situation until recently. What I mean is, the current situation arises by the desire of Western businesses of getting hid of productive investments and concentrating only on capital investments. It has nothing to do with trading with an authoritarian government or not, which almost everyone believed was Ok until recently.
> The graphs may show economic advantage. It’s hard to quantify the long term strategic and militaristic disadvantage to not being able to make anything yourself if a world war occurs.
Is the United States at risk of not being able to make anything ourselves? We have the second largest manufacturing output in the world.
Pricing in externalities (such as national defense impact) is a basic function of economic policy.
I searched 'economics 101 strategic industries' and found this[1] within 30s which includes an overview of 'national self-sufficiency'. It presents the standard argument, including the parts you claim the standard argument ignores.
I personally favor decentralized planning over markets, but I find it unnecessary to slander economics.
It will actually happen because it’s nothing new. The 500b is almost all wages for existing US-based employees. They are looking for a carve out from the new China tariffs (same as last time). Note - they made a very similar announcement 4 years ago https://www.apple.com/newsroom/2021/04/apple-commits-430-bil...
* What type of jobs? - "The 20,000 additional jobs, Apple said, will focus on research and development, silicon engineering and AI."
* Does the US have the required people, in terms of numbers and skills? - "The company is opening up what it calls a manufacturing academy in Detroit, where it will help smaller companies with manufacturing. It already operates an academy for app developers in the city. It’s also doubling its manufacturing fund in the US to $10 billion." - Sounds like they are upskilling, and will count the employees of companies joining the academy as "jobs created"
* Does this mean moving to US based fabs for the M-series chips? - "[M-Series] chips themselves, however, continue to be produced in Taiwan.
* Is this actually profitable, or is this just a political move? - Define profitable. It is cheaper than paying tariffs.
I might be reading it wrong, but that's the 20,000 ADDITIONAL jobs, which is going to be R&D, engineering and "AI".
Those 20,000 people won't be staffing the production lines. So how many manufacturing jobs, especially low skill, entry level with decent pay, will this create? The whole thing is framed in a way that makes it sound like Apple is creating thousands of manufacturing jobs.
> Does the US have the required people, in terms of numbers and skills?
For 30 years, IT managers at blue chip US corporations have exploited the H1-B visa program by saying, "No," and then hiring a never-ending stream of barely-capable Java coders from programmer mills in India, take 5 times longer to make an app than it should have taken, get promoted, and leave everyone holding the bag with shitty web app that we all hate because it's too slow, too bloated, and doesn't work like it needs to. And the companies who can't get enough of that bullshit in-house just hire it out to sub-sub-contractors that do the same thing. Can we not invest in our native population and education systems this time around? I'm so tired of the fact that 90% of the IT staff in my Fortune 250 is Indian, and I know people who would be better at their jobs living in my home town. It hurts our community and our country, in the long run, and by the VERY same logic as re-homing our chip production.
Well, those Indians living in the US will have families of their own, and over time become part of the community you claim to be a part of. Very much like your ancestors did, except they likely didn't face the arbitrary constraints on immigration that Indians (and any other nationality) face today.
The same thing that happened in the UK will happen in America.
People in the UK who are against immigration are often talking about Poles who moved to the UK after the EU and not Indian families who have lived in those neighborhoods for generations.
The crazy thing is, it's not that long ago that Irish and Italian immigrants were not discriminated against. They didn't even consider Italian immigrants to be white.
This need to bend the argument back to the initial English colonization of America is stupid. These mediocre Indian IT drones are not putting everything they own in a boat and washing up here hoping to find a better life. They're the rich B students that could afford the process which become part of an idealized system that American corporations are now bending and exploiting to hire what are essentially indentured servants from a population of people who couldn't get the best jobs in their native country, so they settled on this backup plan.
And they DO have families of their own here (and bring over their in-laws), and a lot of them don't integrate well, for a variety of reasons. At least a third of my neighborhood is Indian. They glare at me on the sidewalk when I wave. And most of them remain inured in their caste system, and are difficult and unpleasant to work with.
Again, all the same arguments about developing our own chips domestically -- which I doubt many people have a problem with -- apply to developing our own, better education pipeline to fully develop domestic software engineers.
It sounds like you should be directing more of your anger to the C-suite than the people they’re hiring. If they couldn’t get even cheaper Indian immigrants you’d be complaining about code boot camp hires instead - what you need is a tech union which would give you the ability to push back against short-sighted decisions which make your life worse cleaning up messes.
> Does this mean moving to US based fabs for the M-series chips?
This is not really a practical option. A big part of the M-series success is TSMC's lead in cutting edge process nodes. And Taiwan does not allow export of technology for the latest nodes. It is available only there.
Developed, sure. Successfully integrated and commercialized, no. Organizations in US and Europe has done a lot of the prerequisite tooling and research. But they haven't successfully integrated it into an operation capable of producing leading nodes, yet.
Do we think Apple will once again sell servers to customers?
I guess they could sell servers to customers who want to run the latest Apple Intelligence models on-prem, even though that probably wouldn't make much of a difference, since you probably still have to trust Apple.
It would make sense for Apple to fork their next highest-end Mac Studio motherboard, make relatively minor changes to it (e.g. add a higher bandwidth NIC and strip out unnecessary I/O) then wrap multiples of those into a rack mount chassis, with commodity-grade cooling and power supply solutions appropriate for the context.
Combined with a properly headless fork of their OS stack (think Darwin, not OS X Server) they could spin up a highly competitive solution using entirely "B-team" resources.
...then it would be piped through their design-council, run through 5 more iterations to get a unique unibody case for it, accompanied by an optional proprietary Apple rack and a price-tag triple of the competition.
That's along the lines of how it usually rolls whenever Apple tries to make something purely utilitarian, it's the most considerate and "fresh" look at a product, but ultimately designed to be used and then disposed when finished.
A purely utilitarian IT-appliance without a individual end-user doesn't seem to be possible in their product pipeline, you usually end up with something "Prosumer": Impressive on its own, yet of degraded maintainability and scalability.
It's like asking Bugatti to design a public transport bus. It would surely be an impressive bus, but not one you would want to maintain over years at a scale of hundreds.
> Apple said that it, together with Foxconn Technology Group, will later this year begin producing the servers that power the cloud component of Apple Intelligence — a system called Private Cloud Compute — in Houston. That marks a relocation, at least for some production, from overseas. Next year, it says a 250,000-square-foot facility for such manufacturing will open in the city.
[...]
> Apple will also expand data center capacity in Arizona, Oregon, Iowa, Nevada and North Carolina, all states with existing Apple capacity. The company confirmed that mass production of chips started at a Taiwan Semiconductor Manufacturing Co. facility in Arizona last month. Bloomberg News recently reported that plant is building chips for some Apple Watches and iPads.
I think they might build a cloud offering. Something like Cloudflare workers but AI centric, perhaps running Swift on the Apple equivalent of V8 isolates.
Makes sense from a business perspective - there's significant growth potential for them as their presence in web tech is approximately nil.
Whoa, whoa, whoa, on occasion, if the planets align just right, I can also get Siri to set a reminder (and at least half the time Siri gets it 80% right).
I’m confident that LLM’s will not have hallucination problems in the type of requests that I send to Siri.
I don’t ask Siri for facts (just like I don’t ask LLM’s for facts). As long as it can correctly, understand what and when I ask to be reminded about something, that would be a huge improvement for me.
That and being able to map “Bedroom Fan”/“Bedroom Fan Light” to “Bedroom Fan Lights” without having to specify aliases (and even then it hearing me wrong).
I’ve see Home Assistant working with LLMs and it can understand groupings that I never explicitly defined which is very nice. I can say “Turn off all overhead lights” and it will find all my overhead lights and turn them off. Siri/Alexa can’t handle those tasks currently.
It's the other way around. The model is impeccable at "understanding text." It's a gigantic mathematical spreadsheet that quantifies meaning. The model probably "understands" better than any human ever could. Running that backwards into producing new text is where it gets hand-wavy & it becomes unclear if the generative algorithms are really progressing on the same track that humans are on, or just some parallel track that diverges or even terminates early.
Only if you wildly oversimply to the level of being misleading.
The precise mechanism LLMs use for reaching their probability distributions is why they are able to pass most undergraduate level exams, whereas the Markov chain projects I made 15-20 years ago were not.
Even as an intermediary, word2vec had to build a space in which the concept of "gender" exists such that "man" -> "woman" ~= "king" -> "queen".
3 lines? That's still going to be oversimplifed to the point of being wrong, but OK.
Make a bunch of neural nets to recognise every concept, the same way you would make them to recognise numbers or letters in handwiting recognition. Glue them together with more neural nets. Put another on the end to turn concepts back into words.
... Oh interesting. And those concepts are hand picked or generated automatically somehow?
> For a less wrong but still introductory summary that still glosses over stuff, about 1.5 hours of 3blue1brown videos
Sorry, my religion forbids me from watching talking heads. I'll have to live with your summary for now. Until I run into someone who condensed those 1.5 hours in text that takes at most 30 min to read...
> Oh interesting. And those concepts are hand picked or generated automatically somehow?
Fully automated.
> Sorry, my religion forbids me from watching talking heads.
What about professional maths communicators who created their own open sourced python library for creating video content and doesn't even show their face on most videos?
You're unlikely to get a better time-quality trade-off on any maths topic than a 3blue1brown video.
He's the kind of presenter that others try to mimic because he's so good at what he does — you may recognise the visuals from elsewhere because of the library he created[0] in order to visualise the topics he was discussing.
Simplifying to that point is more of what a Markov chain is. LLMs are able to generalize a lot more than that, and it's sufficient to "understand text" on a decent level. Even a relatively small model can take, e.g. even this poorly prompted request:
"The user has requested 'remind me to pay my bills 8 PM tomorrow'. The current date is 2025-02-24. Your available commands are 'set_reminder' (time, description), 'set_alarm' (time), 'send_email' (to, subject, content). Respond with the command and its inputs."
And the most likely response will be what the user wanted.
A Markov chain (only using the probabilities of word orders from sentences in its training set) could never output a command that wasn't stitched together from existing ones (i.e. it would always output a valid command name, but if no one had requested a reminder for a date in 2026 before it was trained, it would never output that year). No amount of documents saying "2026 is the year after 2025" would make a Markov chain understand that fact, but LLMs are able to "understand" that.
They are AMAZING at understanding. I think even more reliably than generating. And with context, and back and forth.
Try asking ChatGPT to remember some obscure film you can only remember very hazy details about - really random stuff - I bet it will identify it for you I a few tries.
You can also totally miss spell words, use messed up grammar and it has no problems at all
If you haven't tried OpenAI's advanced voice mode, it's a mind blowing version of exactly what things like Siri really ought to become with a little more development. If that's what you mean by LLM Siri, I totally agree.
Being able to chat casually with low latency, correct yourself, switch languages mid-sentence, incorporate context throughout a back-and-forth conversation etc. turns talking to these kinds of systems from a painful chore into something that can actually add value.
...that will grind your request to set email Vacation Mode through the world's worst speech-to-text, jam the text into Chat GPT, and spend the next three minutes reading you an uninterruptible 3 minute essay about violence.
Over on Android it's the opposite situation. The voice interface to Google Assistant was very reliable for simple things like reminders and appointments, and even for general knowledge questions. It was part of why I didn't switch to an iPhone. Then Gemini came along, and that core functionality got a lot worse.
I tried this with the new ‘apple intelligence’ that I thought could see my screen.
I had a birthday invite with clear date and time: so I asked it to add to my calendar.
It just said, “add what?” repeatedly until finally deciding it needed to send it to chatGPT to help. Which it did, then just returned the text in the image without taking any action. Then I say “can you add the event now?”
“Add what?”
So I try copying the text from the image in photos and giving that to Siri to add as an event. Surly this can work?!?
When Siri first debuted it would automatically beep, so I could immediately tell if the phone did not recognize recognize "Hey Siri" (just "Siri" didn't work). A couple of iOS updates later this went away, which means I can't tell without actually picking up the phone and looking at it whether the command was accepted.
Even more annoyingly, sometimes there is a beep! ¯\_(ツ)_/¯
The problem is AI current best use case is creative work, art, music, programming, but skilled creative professionals is a/the core userbase for Apple products.
Apple is stuck and it’s AI will never be good enough until those creatives embrace it. Right now it’s disdain when mentioned.
An oft-cited quote goes something like this: "we wanted robots/AI to automate boring, routine, meaningless jobs to let people be free to pursue arts, music, creativity. It's a sad state of affairs that AI is taking over arts/music/creativity stranding people with boring, routine, meaningless jobs"
AutoCAD came to the Mac when Intel was shitting the bed (with aggressive OEM contracts for first party system integrators that prevented AMD adoption across HP/Dell/Lenovo-lines) and Windows 11 was being forced on users.
WINTEL played the monopoly game too hard and is starting to lose ground.
I don't know if this is an actual problem you have, but since Siri appears to be composed of independent voice-to-text and text-to-action systems, you can say "start a one three minute timer".
When I put in timers -- for some reason my timer frequently/randomly just sets to 79 hours and a random assortment of minutes and seconds. I have no clue why. I always have to double check otherwise I might be waiting awhile.
It feels like it was a residual timer or something but I have never set anything like that - it is quite strange.
Funny I turned Siri off because i didn't want apple intelligence running amok. The follow-on problem --> lack of Siri killed my Carplay because Siri is required (also use itf for setting alarms/timers). The kicker? I can't seem to turn Siri back on after look through all the menus.
I.e. My preference for apple CarPlay supersedes my concerns on GPT running over my contents. Though the UI/UX has made it next to impossible to turn it back on.
If your phone is new enough for Apple Intelligence, Siri is now under that umbrella. There's no "just Siri" option anymore, unless you're rockin an iPhone 14 or older.
Apple Intelligence and Siri are still separate (though Apple like to make it look like they are fundamentally intertwined). You can turn Apple Intelligence off and leave Siri on for CarPlay.
How did you turn Siri off in the first place? That's where I'd start...
The part of Siri that causes the most trouble is the speech recognition - which uses a voice recognition model that we now colloquially refer to as "AI." The part that works reliably, the part that sets your alarm or sends the message, is an action that's hardcoded.
IMO, moving towards AI just leads to increased uncertainty and undesirable outcomes, which is something several journalists reviewing Apple Intelligence have attested to.
I'm a formerly non-mac guy who finally bought a brand new iPad. I got bored with the wallpaper but couldn't figure out how to change it. "Hey Siri, how do I change the wallpaper?"..."Sorry, I can't help with that". Tried a couple more questions and all it did was Google it for me. This is the latest M4 that was around $2k.
This is what our "AI accelerated" chips give us in return? What a disgrace
We used to call this an "Industrial policy" or an "Economic development policy". Back in the golden era when a strong labor movement coexisted with a Red Scare. 78 years after Taft-Hartley and 44 years since PATCO, not so much.
We have maybe fifty or a hundred million people rotting away in areas where jobs are scarce and housing is plentiful, because we used government policy to shut them out of areas where jobs are plentiful and housing is scarce. We systematically exported jobs from places that aren't big cities because they can be performed overseas and our aristocracy can still profit from them by owning those people overseas.
I don't know if returning to a little more deliberate of an economy is even a partial salve for the place we've found ourselves, but I don't think this laissez faire thing is sustainable for a whole lot longer. We are overleveraged, and arrogantly delusional about our sway at the moment; "Ownership" is not some valuable skill. The fall of an anchor currency and global conversion to an alternate financial network would be a spectacular thing, an astroid striking terrain, which might leave craters on entire other continents from secondary ejecta. World wars have been fought over less.
"We" wasn't big government. It was a million homeowners who decided that the neighborhood they moved into should be frozen in amber forever. Everyone wants housing to be cheap but also for their property values to rise onto infinity. They push back against any attempt to change this and then complain about the inevitable results.
Oh god. Hopefully they’re lighter than the old xserves. We had one still running up until a few years ago when we finally removed it. You could break a toe if it dropped while pulling it out of the rack. People are still selling them on eBay.
I suspect that was the RAID drive bay (ridiculous item). That was a 3-4U monster.
The Xserve, itself, was a 1U unit that was pretty much the same (or lighter than) any other 1U server (we also had HP and Dell servers that were heavy). The weight distribution could be weird.
That stupid drive bay was a proprietary nightmare. The disks cost a fortune.
However, from what I can see, this will be for "internal-use" servers. I don't think they will be selling iron; just services run on the iron.
I'm curious what they will look like, given that these are not for anyone else to buy. Maybe Apple made a different form factor/configuration that suits their datacenters better?
I don't think it's AI servers for Apple to sell. I think it's just regular x86 Linux servers to power Apple's AI cloud services. It's a commitment to internal investment rather than a product.
They have hiring positions for running a Darwin-based server OS, and their private cloud compute is on Apple Silicon. I doubt it's going to be swaths of x86.
Does this mean competing with Asahi to run a Linux kernel, or will this be an attempt to run AI workloads on XNU?
Consider the cost of GPUs, losing what could be double digit percents on overhead might not make this very competitive. The macOS microkernel can still beat NT in some situations (like not having filters slowing filesystem down to a crawl), but it lags significantly behind the investment in Linux performance over the years by every other major company.
I don't think it's AI servers for Apple silicon. I think it's just regular x86 Linux servers to power Apple's AI cloud services. It's a commitment to internal investment rather than a product.
Apple doesn't like B2B and Steve Jobs was very vocal on this (there are various videos where he explains why). Ever since they can afford to, they reduced their B2B to the bare minimum needed. So don't expect anything server-like from Apple.
I'm not sure. Because the moment you enter the system room, the ecosystem is a completely different universe.
You can't easily sell "Good / Better / Best" version of a single model, and tell "These are the options, take it or leave it". Servers are customized to the screws they come with and are expanded throughout over the years. So, the logistics are somewhat different for these kinds of devices.
Plus, macOS is not a CLI first operating system for server operations, and macOS Server is not updated for some years. Allowing Linux would be a different offering, and allowing macOS to work with all kinds of hardware from ordinary Ethernet to 100G+ Ethernet and 400gbps Infiniband (plus all the other interconnects) will be a fun exercise in testing flexibility of both macOS and Apple development teams.
So it's quite complicated. All servers are built to order SKUs. Dell keeps configurations "per server" in their databases, for example. If you have a Dell server, enter its service tag to support site, and you'll get the configuration of the device as it left the factory.
They don't have to go all the way though and fully compete with Dell EMC/HPE. I'm not sure what the original commenter was thinking but in my mind they could simply sell a Mac Studio variant with dual PSUs, better networking, a rackmount chassis, etc. Basically have their existing consumer machine placed in a more datacenter friendly factor.
I mean places like Github and AWS are painfully racking up Mac Minis for their deployments and this theoretical server model would simplify everything. It also becomes an option for on premises AI inference using MLX, especially if they manage to get ANE support working in conjunction with the GPU for faster prefill.
The support and software stack for the server model would be the exact same as the consumer variant and they certainly wouldn't have special Linux offerings, Infiniband, and all that. If there's networking beyond their existing 10G it's going to be built into the board and they aren't going to support random 3rd party cards. The unit also doesn't need to be upgradable either.
I believe the intention is to use their own M-Series CPUs - to get what they call "Private Cloud Compute". The cpu on your phone will encrypt data and a request, send it over the network to am M-series CPU which will decrypt and process/send back an encrypted response.
The idea being there's no VMware, kernel or piece of hardware that can have backdoors built into unless someone files off the top of the chip and somehow probes the silicon
> Apple said that it, together with Foxconn Technology Group, will later this year begin producing the servers that power the cloud component of Apple Intelligence — a system called Private Cloud Compute — in Houston. That marks a relocation, at least for some production, from overseas. Next year, it says a 250,000-square-foot facility for such manufacturing will open in the city.
> The Private Cloud Compute servers use advanced M-series chips already found in the company’s Mac computers. Those chips themselves, however, continue to be produced in Taiwan.
IMO we've really got to start pressuring our own governments to take control of their networks, as well as the companies the population is going through (not just Facebook, but even international contractors for services). Letting a foreign government have this much control over the data of your populace and the ability to feed whatever algorithmic message they like is a path bound for disaster in the long run. The powers of the world are way too consolidated as is, and a company can turn into a state actor at the drop of a hat. I don't think we can maintain kayfabe about the country/corporate divide. I also think this can be done without impacting freedom of speech for your population, as long as you don't consider corporations people.
Most countries don't have the resources to do much, but even then they can try their hardest not to be beholden to any single foreign country coughChina, Americacough.
Anyone keeping count of how many trillions in hypothetical investments and millions of jobs large American corporations have promised in the next 3-5 years?
Reverse technology transfer from countries like China is kind of fair. But EU companies should be very wary of Trump tariff blackmail that forces them to build production lines in the US.
>But EU companies should be very wary of Trump tariff blackmail that forces them to build production lines in the US.
That's the whole point of tariffs, to encourage domestic production.
Put another way, what is the difference between what you wrote and
>But US companies should be very wary of EU tariff blackmail that forces them to build production lines in the EU.
?
Whether a particular tariff is economically viable is a reasonable debate. Calling Trump's tariffs "blackmail" without assigning that epithet to all tariffs from whatever source is not.
Not a lot of them, but Trump accuse EU of tariffs because it has VAT. That’s not the same thing because VAT apply to domestic companies as well so there’s no real unfair advantage.
You could make the claim that it’s still worse for foreign goods as VAT can be offset with VAT from supplier purchases, but that only works if you produce goods in the same country as you sell them. But US companies have even larger advantage already as they don’t have any VAT, so in the end the playing field is kind of level.
This doesn’t stop Trump from pressuring and squeezing and making false claims to justify his standing point.
Expect an unstable trade environment due to tariffs and retaliation. This will hurt small businesses that have even more complex environment to operate in.
There's a lot of understandable skepticism about this announcement because there are so many PR announcements. I want to temper that with an alternative perspective.
I actually think Apple is a potential dark horse when it comes to AI hardware. What we have now is essentially a three-layered monopoly: ASML, TSMC, NVidia. This has been incredibly lucrative for NVidia. But, as we know, Apple doesn't like to rely on third-party hardware. They've invested heavily in ARM going back to buying PA Semi [1]. Apple replaced Intel chips (which originally replaced Power chips) with the M series in recent years. Apple is in the process of replacing Qualcomm modems in iPhones, which is not only a technical feat but a legal one given Qualcomm's patent dominance over 4G/5G.
Apple has the resolve for long-term initiatives that few other companies have. Apple Pay continues to chip away and get slowly better in a way that, if it were a Google product, would've been cancelled, rebranded, relaunched probably 3-4 times by now (Google Checkout, Google Wallet, Google Pay, Android Pay, etc).
Apple clearly sees AI as a strategic issue. They have loads of cash on hand to finance basically anything they want. And they won't want to be beholden to NVidia.
I expect Apple to have a significant impact here but it won't be tomorrow or even this year. It'll be over the next 5-10 years.
Apple's incentives have definitely aligned with replacing Nvidia entirely ever since they ceased diplomatic relations. But Nvidia also knows this, which is why they invest heavily in things Apple will never do. They write the official Linux drivers Apple wouldn't get caught dead supporting. They give users and integrators freedom to choose their OS, software and library stacks to better suit their application. They sell individual GPUs and unlocked edge compute hardware with no distribution terms or $99/year "developer license" bullshit. Nvidia is a hardware company in places where Apple tries shipping services instead.
Then there's also the software issues. Nvidia has invested in GPU-based compute nonstop for the past 10+ years. Apple invested in Nvidia, then invested in OpenCL after abandoning Nvidia, then abandoned OpenCL for Metal compute which would eventually become the proprietary Accelerate framework. Nvidia's eggs are all organized in one, valuable basket. Apple's investments are spread out all over the place, with much of the time and money going into projects that don't even exist anymore.
Apple has the TSMC advantage, but that's just about it. Their GPU designs aren't comparably efficient or compute-oriented to what Nvidia ships today. Additionally, Nvidia will continue investing in places that Apple principally refuses to support. Unless a serious tide change occurs at Apple, they aren't going to get a fair competition with Nvidia.
There's a lot of noise I can see behind the scenes on investor confidence. Noise as in "everything is fucked" sort of level of noise. Thus I expect this is being said to try and stop the AAPL stock collapsing in the upcoming recession that the analysts are predicting more than a tangible expansion and recruitment goal.
I also take issue with their being 20,000 people on the market who are still able to contribute something useful. They will be culled quickly and quietly down the line in the annual corporate lay offs.
It is not the time to make grand gestures unless you're trying to gain political favour, at which point any respect I have at least is gone.
Coincidentally, construction isn’t set to start until late November 2028—convenient timing. If this mess blows over, they can quietly backpedal and carry on like nothing happened.
Good timing because Trump should be significantly weaker, and it'll be clear where Trumpism is headed in the culture, but also it will be more clear where AI will end up.
Even if the move forward with investment, they will be a bit of a 'late' mover, but will have had a chance to see what is working and what isn't working for everyone else.
1. JD Vance independent of Trump will have the same policies.
2. JD Vance will have enough popularity for a serious 2028 run. He might fall out favour with Trump as Trump tries to mount a bid for a third term, Trumpism might just generally lose popularity if policies lead to bad outcomes.
3. Dems don't figure their shit out. They should be able to take back some control in mid-terms, and then start to push their own policies, or at least credibly show that the most extreme policies from the exec branch don't have teeth anymore.
And you're assuming that the 2028 election (and 2026 for that matter) will be business as usual elections, against all evidence to the contrary staring us directly in the face.
The federal government doesn't control elections in the US so they don't have much power here. Also, firing all the FBI agents is a bad first step to using them for interference. They have no clue how to be authoritarians; to do that, you need to be popular and have the security forces like you.
The playbook is obvious - if Trump loses, Vance refuses to certify the election results due to "fraud", Republican states will produce alternate lists of electors that vote for Trump, and he will claim that in reality he just won reelection with the biggest margin in history.
How do we know this isn't just crazy conspiracy theory? Because they already did attempt the same thing in 2020, and this time they had the chance to vet the VP candidate for this scenario.
Then we can have a color revolution. He's not doing the right steps to prevent that, because he's annoying the security forces instead of supporting them.
Parties in the US aren't real, though the brand value is good at keeping the right kind of people in each one. They can't actually kick anyone out and people in the federal government have no particular control over same party people in any state.
From my view, the two parties are quite a bit more real than the 3 branches of government. The people funding state and federal elections are the same.
If you mean campaign funding, that's mostly Actblue (upper middle class Dem donors). Republican donors are like, five people, yes - but voters have their own opinions and they definitely don't vote for whoever has the most money. Because the Democrats have more of it.
I just checked. US Constitution has amendment #22 that says: "No person shall be elected to the office of the President more than twice". Is there a workaround? I doubt he can change that amendment. None have been ratified in the US for a squillion years.
Can they realistically cancel their construction contract right before the construction starts? Sounds implausible to me, at least not without huge compensatory fees
This is an insane amount of cash to throw at any problem. They could have Apple rockets mining asteroids with $500,000,000,000. There is no way all this cash goes into AI. What will actually happen is they will take 1/10th this cash to an over-valued startup and acquire them.
Between land, hardware, having to build multiple power plants, the labor costs involved in all of this, and setting money aside to run all of those for however long, then yeah I can see where it gets up to that price. 20k engineers is easily 5 billion per year in salaries, probably more.
Assume the human resources at $100,000 per head, and you get $2B/yr. Four years comes to just $8B for human resources. Assume land costs $10B. Assume construction costs $100B.
653 comments
[ 3.4 ms ] story [ 429 ms ] threadApple's obsessive focus with raster efficiency really shot their GPU designs in the foot. It will be interesting to see if they adopt Nvidia-style designs or spend more time trying to force NPU hardware to work.
That said, an M4 Ultra (extrapolating from Max and Pro) would likely compete with my 3090, and with 192GB of memory (for 10x the amount it should cost) will out perform my 3x3090 AI server. And honestly, cost less than my 3 3090s + rest of the computer + electricity.
It won't outperform a bunch of A/H 100s (or even a single one, or any other cards in the enterprise realm) though, but it will cost an order of magnitude less than a single card.
inference is not the same as training.
https://security.apple.com/blog/private-cloud-compute/
Maybe instead of saying the t-word tariff, US gov can charge Apple a special fee on each iPhone. They can call it something catchy, like say, a Core Technology Fee.
https://edition.cnn.com/2019/09/23/tech/apple-mac-pro-united...
Measurable results affect rational aspects of our minds, PR attempts to attacks directly emotions bypassing the former, ie to induce impulsive shopping.
Also, what actual security? Apple is as vulnerable as cheap chinese phones against state actors using 0days. Apple devices are still being stolen for spare parts, Apple doesnt secure each component AFAIK and thieves know this (very recent case with friend of a friend, they even knew how to bypass that built in airtag tracking). I haven't seen anything but very well crafted PR statements on this topic. All money-accessing apps on absolutely any phone are a security risk.
But folks love convenience above all.
Sounds more like technical marketing and the company will treat any decisions around it as a marketing exercise.
Not entirely true. Some of the highest value components in an iPhone, including the CPU/SoC, baseband, and the majority of OLED displays, are sourced from countries that are not mainland China.
> and mostly in the same geographic region
Most smartphone supply-chain for Samsung and Apple exist outside China -- primarily in Japan (camera, sensors), South Korea (DRAM/NAND, OLED), and the US (various ICs fabbed at TSMC in Taiwan). There are quite a few reliable estimates/teardowns showing that these three countries account for close to about 90% of iPhone BOM (bill of materials). That's one reason why Samsung's smartphone unit was able to pull out of China without much disruption back in 2019 -- ie, low dependence on China.
I feel that Apple has pushed this misleading narrative a bit too long to defend their massive China outsourcing.
Apple could, with its immense cash hoard and cash flow, _make_ the US viable, but it chooses not to because it'd rather take the easy way out and have China or India or $COUNTRY fund it and return money to shareholders. They've returned money to shareholders rather than invest it in US operations, by design.
This is a classic feint to protect Tim Cook's entire raison detre. He built his career on super high efficiency operations by outsourcing to cheap labor countries. It relies on the low-to-no tariff access to US consumer money.
And I don't care that it's better for their stock price; that's Apple's problem not mine as a US citizen. And even as an Apple investor I would rather the money be spent on US on-shore operations.
> And I don't care that it's better for their stock price; that's Apple's problem not mine as a US citizen.
That is the shareholder’s problems. People like to think that their investments won’t go batsh*t insane overnight.
> And even as an Apple investor I would rather the money be spent on US on-shore operations.
Apple doesn’t even make all, or even most of its money in the states. Not all of its shareholders are American, if they went this route they could lose half of their revenue overnight (as other countries note the protectionism and tariff or simply forbid Apple products from being sold).
>Apple doesn’t even make all, or even most of its money in the states. Not all of its shareholders are American, if they went this route they could lose half of their revenue overnight (as other countries note the protectionism and tariff or simply forbid Apple products from being sold).
So? I don't care about their revenue, I care about the future of American industry. Having a bunch of cash hoarded by old people is irrelevant if it isn't reinvested in something I care about. And I don't care about your supercar or Nobu reservation, or if some fund returns an extra 2%. This is despite being a direct beneficiary.
Live by shareholder return, die by shareholder return; the US is not and shouldn't ever be geared to shareholder return over everything else. Apple and other companies have freeloaded off the US for far too long.
Juche doesn't work in North Korea, it isn't going to work in the USA.
If the Trump administration has any competence, they will rub those old promises in Apple's face until Cook actually does something meaningful.
And Dell became a case study of outsourcing everything (and sending your stock and profits soaring the whole time), until you have nothing.
Ask maybe China.
It destroyed the futures of a larger number of Americans.
Then again, why do we make the distinction "American"? If you have people who became unfathomably wealthy by shipping off strategic industries to the lowest bidder regardless of geopolitical implications, does nationality matter anymore?
Way easier and more globally optimal than just saying we're going to do absolutely everything (even the shitty jobs) here in the US.
W-2 get refunds because the Feds took out too much from their paycheck beforehand.
You can choose to withhold more or less, but the default taxation on w-2's do generally give a bit of a refund. Better to take out too much when you don't need it than slam down a gigantic bill when at once a year later.
Yes.
The on-average crossover between negative and positive net total federal income (individuals will differ because of individual circumstance beyond just income level) tax when taking into account refundable credits (most notably, but not exclusively, EITC) is a bit below the median personal income but not that far below it, so certainly lots of individual "normal" (by most reasonable definitions) Americans do not pay net federal income tax .
But even if they don't pay net federal income taxes, they probably still pay a net positive amount in a variety of state taxes, federal payroll taxes, and federal consumption taxes (e.g., gas tax.)
https://www.newsweek.com/nearly-all-republicans-vote-against...
If we don't put pressure, those people will get their way.
To be clear, I do think capital gains taxes are criminally low in the US relative to income tax, so I'm not arguing in _favor_ of lower taxes. I'm just saying why raising taxes isn't a panacea.
You’re not wrong, of course, about how every tax percentage point matters. But Americans arguing that their taxes are too high is never not hilarious.
So make the US to be like a far less successful country? Kill your economy by increasing taxes? The US economy is singularly successful because it has incentives to build businesses - see YC.
Have you tried living in a country that doesn't encourage businesses? They are often great tourist destinations. I'm in New Zealand and too many ambitious young people leave here: we have an emigration problem because our economy sucks. The government fixes the economy with 30% immigrants (disclaimer: I love immigrants). I have many friends that are never coming back here except for holidays. I hate the New Zealand government incentives for businesses (taxation and regulation) and I can see no way to fix them. Even our "business" political party ACT is completely fucked (latest story - they will be selling everything profitable to overseas "investors" - destroying the economy).
Taxation incentives matter to businesses. Be careful what you ask for because the majority have little understanding and vote for the wrong incentives.
Even business owners don't seem to understand incentive systems that well. Perhaps game designers do?
Then why would ensuring the same effective tax rate on the 99th percentile kill their incentive?
However I believe that incentives need to be marginal. If you already have a lot perhaps you need a big carrot as your incentive? I don't know any billionaires that I can ask how they feel about taxation incentives: I reckon you are making assumptions about what you think they should feel.
What makes Tim Cook make the US more money?
Taxation cliffs are shit. In New Zealand our Green party decided that 1 million was enough. Why would you bother growing a business after you reached 1 million? Retirement? A business is defined as being about making money (albeit some people do run "businesses" for other outcomes - why is Warren Buffett still working?).
High marginal taxation is also shit IMHO.
The hard part is to design the incentives so that productive people build your economy for the benefit of everybody.
If a government discourages business then the economy is crap and everybody suffers. See other economies.
Few people understand the incentives of others, and few people understand how wealth is created for all: the hoi polloi dismiss the wealthy as vampiric money grubbers. Anyone who uses the word capitalist in a derogatory way has been brainwashed. Most everything that makes our economies work is invisible non-monetary rewards. https://news.ycombinator.com/item?id=43162596
I can speak for my own financial incentives. My perception is that I have an effective tax rate of well over 50% in New Zealand (any retirement savings are not safe because our demographics and governments will screw our economy).
I do not feel the incentive to work in a business - My attitude means I now produce marginally less than I could for the New Zealand economy (I still pay taxes so they are advantaged but they could get a lot lot more from me). I now mostly selfishly concentrate on those closest to me. Why should I work if it isn't marginally beneficial enough for me? I'm no more selfish than my retired friends that I know (a wide variety of people from many walks of life).
(Reëdited to expand and clarify).
We can't decide how much is fair. Compare yourself to a dead king - what is fair? We can design systemic incentives so that we each make the world better for everyone. Not that that it is easy... Trite thoughtless dismissals of the most productive members of society are not helpful.
Edit 2: I guess this discussion is as close to work as it gets for me. Too much adulting. Should I get into politics? Are morals an impediment to helping others? There are too few politicians I admire, and too many I wouldn't want to shake hands with or be associated with. Every idiot has political opinions - how much of an idiot am I? Every politician is smart enough to win an election - they are not stupid yet they make too many horrific mistakes. What about the cryptically smart ones? I see how systems affect people that join a system. What would I become if I join our political system? Understanding our different systems is hard because they grow so weirdly with vestigial complexities due to history, complex interactions, and reflexivity.
They are parasites at this point. If they think they can find lower taxes than 22% they are happy to leave. As if they aren't already avoiding taxes.
Do Medicare and Medicaid exist without businesses? I'm from New Zealand and our society causes problems for our socialised healthcare.
Businesses are symbionts: productive societies accept some costs from businesses so long as the society get more gains.
Why do you look at money as though that is all that matters?
Who measures the benefits we get from modern society?
Finding downsides and complaining about them is too easy. Looking for upsides is less popular.
Every poor person I've met avoids taxes.
In a purely technical sense, yes. Because you don't necessarily need an American salary to pay taxes that cover these facilities.
It was very much a concerted effort for most other non-govt Healthcare to be tied to often American jobs. Which of course causes a cacophony of problems when less employers are even offering full time work.
>Why do you look at money as though that is all that matters?
It does not, but business these days sucked up enough money that it's starting to affect basic survival, let alone any pursuit of happiness. There's no point finding upsides when the common person is is so low on the totem pole.
Making hypotheticals of "well look on the bright side, you're not dead" doesn't help either. When America starts using that wealth to make sure no one in a first person country isn't dying on the street, we can discuss the subtleties of capitalism.
>Every poor person I've met avoids taxes.
Well I can't speak for New Zealand. You can't tax a poor person with no income. That's how bad the situation is here.
That is a weirdly employee centric view. I'm talking about the US economy. American salaries depend on American businesses. America has some of the best healthcare available in the world. If US businesses are fucked due to the beliefs of citizens (or whatever else), then the US socialised healthcare is fucked too. There's plenty of poorly run countries to compare against (including Cuba where I discovered their lies about their healthcare first-hand as a tourist). NZ socialised healthcare is okay but our economy is not improving and regardless of our desires for more, the social benefits have no choice but to match our economic output.
> it's starting to affect basic survival, let alone any pursuit of happiness
Only if you're one-eyed. US citizens are the rich. In a fair world we would tax all Americans at 90% and redistribute that to the poor in the rest of the world. Maybe same for NZ too (Wikipedia shows that NZ's disposable median income is ⅔ that of the US however it also strangly says that NZ's median wealth is nearly double that of the US -- I'm guessing because houses are more unaffordable in NZ). Income is usually a better measure within an economy of useful output (economies can't really save for next year). The US federal poverty line is about $16000 for one person - a hell of a lot of money for people in many countries.
> Making hypotheticals of "well look on the bright side, you're not dead" doesn't help either.
I guess you're referring to my comment "Compare yourself to a dead king - what is fair?".
My obfuscated point is that few people (maybe narcissists) would give up their modern life to live in past poverty. Antibiotics, freedom, technology, access to the intellectual output of the world. We are mostly a lot better off than the past. Most people don't value that instead they are money-centric (as many of your comments are). Most people seem to compare themselves to people that are wealthier than themselves and then complain about how they are not getting their fair share. Few people compare themselves against the global poor and then talk about how much they should share their wealth downwards. They talk about how others should share their wealth - they rarely seem to consider how they should share their own wealth. Especially ironic given that it appears that the majority of commenters on HN are the wealthy of the world (and often part of the tech overlords - e.g. YC).
The US is often a parasite upon other countries. If you were to say that the US pays it back to poor countries with technology (mostly from rich companies), then you would be implicitly arguing that wealthy US companies deserve to be wealthy. I recall that weapons are the biggest US export (nice!)
I guess I'm saying is really take care not to kill your geese laying golden eggs (even if you think the geese seem to be keeping too much golden egg to themselves): the socialised good that you have depends on those geese (US businesses). The bad is bad but don't destroy the good.
An economy is a delicate balance - as shown by many failed economies.
> When America starts using that wealth to make sure no one in a first person country isn't dying on the street
I've think I've answered that here: https://news.ycombinator.com/item?id=43148513
Cheers
https://en.wikipedia.org/wiki/List_of_countries_by_wealth_pe...
https://en.wikipedia.org/wiki/Median_income
* People want to feel like their preferred lifestyle and policies have no negative implications at all and they don’t have to feel guilty about them.
* People want to feel like they’re part of a group of special people poised to change the world, and everyone else is hidebound bigots who resist temporarily but will eventually be forced to recognize their genius. People want to virtue-signal: demonstrate that they have the good qualities that their ingroup considers most important.
* But people also want to vice-signal: demonstrate their willingness to breezily dismiss the supposedly good qualities that the outgroup considers important.
Was the New Deal an industry-destroying handout?
https://www.pbs.org/newshour/economy/making-sense/steel-tari...
Nor can they move to these offshore places (where the cost of living is lower) because immigration laws exist in part to control worker mobility.
Compared to now where there's just no full time jobs and also wage suppression?
That's the talking point, but it's bullshit. A lot of those "low-value industries" are fundamental capabilities, and China sure as hell isn't going to let the US own the "higher value-add" areas. They dominate those next, and the US free-trade business elites will be fine with it as long as they get to make some money for themselves.
China's going to put the money into making GPUs, and they're going to get it right, probably sooner than later. Then they'll drive the American manufacturers out of business, like they've done in many areas before. Their government isn't beholden to the profit-focused capitalist attitude that is one of the West's biggest vulnerabilities.
Also, it's pretty foolish to 1) forget power and security doesn't come from rarefied high value-add stuff, 2) that there are a lot of people that can't be employed doing stuff like making GPUs.
No. I'm not sure where you got that idea. If you look at something like the Boston Tea Party, it wasn't high taxes on tea that were being protested against, it was lowered taxes on tea that undercut the smuggling operations of people like Sam Adams and John Hancock. "No taxation without representation" makes better press than "No undercutting my smuggling operation" though.
In the early years of the US, between 80 and 90 percent of federal revenue came from tariffs. Not exactly free trade.
To be fair, the Federal Budget back then was 2%-ish of GDP. And their political consensus gave the Federal Gov't very few things that it had the power to tax.
https://pt-m-wikipedia-org.translate.goog/wiki/Quinto_do_our...
Important to note that outlays do not equal federal taxes, because spending exceeds revenue by a substantial margin.
Is a better graph
There’s been an isolationist wing in tech as long as I’ve been in it (early 2000s). I remember chatting with someone at Cisco/Juniper in the late aughts about Huawei ripping off their router designs down to the silk screening. Of course today Huawei makes their own state of the art routers with their own silicon, and some lower-end Cisco/Juniper gear is white boxed foreign equipment. And of course tech folks were complaining about immigration and outsourcing back in the early 2000s when Republicans were enthusiastically supporting both.
The great rule of conduct for us in regard to foreign nations is, in extending our commercial relations to have with them as little political connection as possible. So far as we have already formed engagements let them be fulfilled with perfect good faith. Here let us stop.
Europe has a set of primary interests which to us have none or a very remote relation. Hence she must be engaged in frequent controversies, the causes of which are essentially foreign to our concerns. Hence, therefore, it must be unwise in us to implicate ourselves by artificial ties in the ordinary vicissitudes of her politics or the ordinary combinations and collisions of her friendships or enmities.
Our detached and distant situation invites and enables us to pursue a different course. If we remain one people, under an efficient government, the period is not far off when we may defy material injury from external annoyance; when we may take such an attitude as will cause the neutrality we may at any time resolve upon to be scrupulously respected; when belligerent nations, under the impossibility of making acquisitions upon us, will not lightly hazard the giving us provocation; when we may choose peace or war, as our interest, guided by justice, shall counsel.
Why forego the advantages of so peculiar a situation? Why quit our own to stand upon foreign ground? Why, by interweaving our destiny with that of any part of Europe, entangle our peace and prosperity in the toils of European ambition, rivalship, interest, humor, or caprice?
It is our true policy to steer clear of permanent alliances with any portion of the foreign world, so far, I mean, as we are now at liberty to do it; for let me not be understood as capable of patronizing infidelity to existing engagements. I hold the maxim no less applicable to public than to private affairs that honesty is always the best policy. I repeat, therefore, let those engagements be observed in their genuine sense. But in my opinion it is unnecessary and would be unwise to extend them.
Ironic that as a Canadian, the US is moving from the nation that would be guided by Justice into the belligerent nation in this situation.
It also serves as a lesson to us that we should have learned from you and George Washington, and stood on our own first and ensured our own security before cooperating with others. We have a long way to go to get back there now, unfortunately under the position of potentially our closest ally and economic partner being belligerent, untrustworthy and unreliable.
I believe the body of people who are promoting such a narrative (and it’s clearly coordinated) have major conflicts of interests, they want to maintain the status quo for their own enrichment at the expense of any specific nation, because in reality plain old paying attention shows nearly the exact opposite is true in every sense.
They are very good at their propaganda which is exactly how they got in that position but they are not looking out for your interests in the slightest, they just want to “manufacture consent” for the forward march of the global hegemony, that oligarchy… nobody in their right mind, with eyes unclouded by hate, would come to these conclusions naturally.
It’s probably the scariest thing too, and it’s nothing new! Go read Chomsky’s book by the same name “manufacturing consent” and he lays out many examples that were happening in the 70s-80s, and they are following the same playbook today just with Ukraine and Gaza instead of Colombian Jungles and Vietnam.
I'm a Canadian. Operating SOLELY on the actions and statements of your executive branch, not the media's reporting but the wording of the government, their executive orders, their direct public statements. your government is increasingly belligerent, untrustworthy and unreliable.
Belligerant -> Constantly making subtle threads of annexation. Calling the Prime Minister of Canada the "Governor" of Canada. Constantly lying about trade deficits that are surpluses and drug and migrant problems that are actually a bigger problem moving north across the border than south.
Untrustworthy -> After renegotiating NAFTA to USMCA and hailing that as a great agreement, now its a shit agreement and he's putting tarrifs on to get more from Canada under the threat (and likely actuality) of causing economic harm to both our countries.
Unreliable -> You were our biggest and staunchest ally. Electing an administration that is actively hostile to our government and sovereignty means you are no longer reliable as an ally.
> they want to maintain the status quo for their own enrichment at the expense of any specific nation, because in reality plain old paying attention shows nearly the exact opposite is true in every sense.
The current actions of the administration are destroying hundreds of thousands of lives in order to enrich a few thousand at the top. They are alienating ally's, destabilizing peace and almost guaranteeing more conflict and war in the near future.
Chomsky warned of the damage a demagogue supported by a disenfranchised, hurting and angry populace can have on a country. This is happening now, and half the country is burying their head in the sand claiming its propaganda.
This is total nonsense, nobody thinks they can annex Canada…
Trump defies typical expectations and does things everyone else thinks would never happen. Often because to do so would be very damaging and idiotic, but that never actually stops Trump.
However you do NOT threaten to annex, take over or otherwise threaten the sovereignty of an ally or friend. That is geo-politics 101, so regardless of how serious he is about doing it, the act of threatening it is belligerent and shows he isn't a reliable trading partner or ally.
Outside of annexing, the trade war is also actively hostile on top of being perpetuated on complete lies.
>they just want to “manufacture consent” for the forward march of the global hegemony, that oligarchy… nobody in their right mind, with eyes unclouded by hate, would come to these conclusions naturally.
The oligarchy has taken over. The few that benefit from the "global hegemony" you refer to, which is largely the 'interests of the rich' are now completely in control of the US.
The US administration absolutely is not looking out for the interests of the average Americans. Most of what they are doing directly hurts most or all Americans in one way or another, and the few things they do that help Americans benefit the rich the most.
Now is the time much of the US populace has always claimed it would stand up against tyranny, an oppressive "lord" class and kings. They are watching it and cheering it on instead.
I’m on mobile but Googling for “Apple tariff waiver” and “Apple tariff exemption” will point you to several news items.
From the 45 presidency: https://www.marketwatch.com/story/apple-escaped-tariffs-last...
Now: https://9to5mac.com/2024/12/17/analyst-trump-will-waive-appl...
It is corruption when president outsources and offshores US jobs, though
US consumers, that’s all of you, are being hammered with taxes on imported goods most of which can’t realistically be produced in the US anyway, to solve a problem you don’t have.
A commitment like this takes years to plan. It can’t possibly be a response to tariffs announced weeks ago. This is all optics.
So it's up to the gig workers if they think they're employed or not. Presumably this depends on how often they do it.
A lot of fake employment and low productivity jobs are in the government/NGO sector, paper pushers, DEI jobs, law/compliance type jobs - that should have been manufacturing jobs instead.
USA has no shipyards and infrastructure is crumbling precisely because of misallocation of resources and labor
That's because of the Jones Act and other poorly designed protectionism.
Heck, the _whole_ Federal government workforce is less than 2% of the total workforce.
> USA has no shipyards and infrastructure is crumbling precisely because of misallocation of resources and labor
Yes, and you have tariffs to thank for this. The Jones Act requires US-built ships for any ship traffic within the US waters.
So the shipbuilders were insulated from competition and just degraded into Defense contract moochers.
Jones Act is the reason US has at least some form of domestic shipbuilding
For example, the Washington state needs new ferries, and their cost is literally almost 10x than the cost for the similar ferries produced in Turkey for the nearby Vancouver, BC. With this kind of cost gap, there is simply nobody who would buy the US ships unless they _have_ to.
Without the Jones Act, the shipbuilders would have adapted long ago to produce high-margin high-tech components for the ships for the international market. And likely in larger quantities than they do currently.
And I don't trust unemployment in this gig economy. I'm technically not unemployed, but I haven't had a full time job in nearly 2 years.
But nothing Trump is doing is going to fix your situation.
In fact he (or rather Elon/doge) is very actively making things worse for you with the massive government layoffs, flooding the market with even more people to compete with you for jobs making finding work more difficult and also eventually dropping all of our wages.
I'm aware. I'm sure he's responsible for at least 3 job freezes I ran into mid-interview this year. He's literally costing me job opportunities because no one can budget around this chaotic government.
In February 2020 it was 63.3% and in January 2025 it was 62.6%, for a difference of 0.7%. Also note the steady decline post-2008 and the multi-year plateau that jitters around 63%.
Having the plateau change from ~63% to ~62.5% isn't an unreasonable scenario.
* https://fred.stlouisfed.org/series/CIVPART
That's because Americans got /richer/ and retired. You want to look at the prime-age rate, which the boomers have moved out of.
If you look at the prime-age employment rate, we're almost up to the record high levels: https://fred.stlouisfed.org/series/LNU01300060
Trump is using extremely misguided legislation from the 1960s/70s where Congress allowed the President to enact tariffs for national security and emergencies. There is a very strong argument (in the sense it resonates with the conservative SCOTUS majority) that Congress cannot delegate its fundamental powers to the executive by legislation alone.
I think people are just too cowardly to bring a case in front of the courts to challenge the constitutionality of it all. Non-delegation doctrine is what the Federal Society want to use to kneecap all federal regulation. Trump operates on a spoils system so it's not in the interest of conservatives or businesses to challenge him, for fear of retribution.
Previously they were outsourcing and offshoring as much as they could get away with it. Which led to transfer of advanced technologies outside USA and America losing its manufacturing and technology edge
I haven't seen a citation from either side. Can you substantiate your position?
TL;DR read Article I section 8, read up about the Trade Expansion act of 1962 and Trade act of 1974, and "non-delegation doctrine", and you can trivially find legal debate about the constitutionality of IEEPA. Rather than listen to random nerds on HN you should seek out this information yourself.
https://www.axios.com/2025/02/20/commerce-secretary-lutnick-...
And both sides of the aisle benefit from this. Whether it's legalized insider trading or jumping to corporate jobs when out of office, it's a corrupting influence. All members of Congress, SCOTUS and POTUS should have to place their assets into blind trusts. That won't stop this corrosive influence, but it is the bare minimum.
The feds can and do go after people for using family members and friends to execute trades. So if Nancy Pelosi told her husband some material non-public info, and Paul Pelosi traded on it, that would still be insider trading.
There was a guy at Microsoft who was caught once using a friend to place trades. He said he talked himself past his ethical concerns by reasoning that members of Congress do it.
https://www.businessinsider.com/sec-alleges-insider-trading-...
Edit: to be clear, the absence of a prosecution does not mean that the Pelosis did not insider trade. Nor that they did. We can't tell from this distance, only speculate.
https://finance.yahoo.com/news/deceptive-tactic-nancy-pelosi...
Oh yes, you can:
"Paul Pelosi, 83, sold 30,000 shares of Google (GOOGL) stock in December 2022, just one month before the tech giant was sued over alleged antitrust violations." https://finance.yahoo.com/news/deceptive-tactic-nancy-pelosi...
https://www.newsweek.com/nearly-all-republicans-vote-against...
The Reciprocal Trade Agreements Act by Congress gave FDR sweeping authority over tariffs without direct congressional approval.
Then later the Trade Expansion Act under Kennedy.
Historically, these have served to decrease tarrifs.
Which means those who care about this are back to not only contacting legislators but also persuading a lot fellow voters that separation of powers is crucial and worth prioritizing over familiar well-handled and loved heuristics.
Congress has become increasingly unproductive and unresponsive. There are many popular policies that Congress essentially ignores, and many problems that go unsolved. So trust in government dwindles and people crave strongman solutions.
I’m not sure there is a solution. There are so many interlocking problems gumming up the process that any “we just need to fix X” solutions (where X is gerrymandering, money in elections, lobbying, the two party system, first past the post, corruption, income inequality, the electoral college, the slow death of journalism, consolidation of industries, etc) are nearly impossible and also probably insufficient because they all feed back into one another: they are both causes and effects.
So when people are mad about a downstream effect like the price of eggs and digging any deeper touches one of the topics above (“to fix egg price gouging you need to reinvent the political system” sounds a lot like “to make an omelette first you need to create the universe”), it’s really easy to throw your hands up.
1. The agents of political change are themselves subject to politics. (Unlike maintainers of a codebase, who are able to make top-down decisions about the code.) If you make a bad decision, it may be the last one you’re allowed to make.
2. There are no unit tests for politics. There is only history, which is an imperfect predictor.
3. Code is deterministic. Policy is not.
1. Politics does enter into programming teams. There's several articles of this genre. Where the most highly regarded programmer made bad decisions over a long period of time & exhibited hubris in doing so. This programmer was fired, but if he was a founder otherwise held some leverage, then he may not be fire-able. However, other staff can leave. It's more difficult to leave the impact of politics.
https://www.freecodecamp.org/news/we-fired-our-top-talent-be...
Entire industries can make bad decisions due to incentives. Such as I must learn technology X because companies hire for technology X. Companies choose technology X because developers know technology X. Technology X could have fundamentally flaws. Compounding complexity in the ecosystem as additional tech is made to fix the flaws...which will also have flaws.
These cycles of compounding bad strategic decisions yet good tactical decisions can last decades.
2. In some ways there are unit tests for politics. Geopolitical signals can be sent. So if X happens, there's a tacit agreement that Y will be the response. Unit tests are great for small units. However, complex systems can be difficult to comprehensively test. This is why there is fuzzing...which is effectively a monte carlo simulation. Monte Carlo simulations are widely used in political analysis.
3. There's non-deterministic inputs. And the system can be more complex than what's possible to model. Leading effectively to non-determinism in making decisions about how to modify the system.
Congress seems pretty unwilling to do much of anything right now.
https://www.cbc.ca/news/politics/rcmp-border-patrol-canada-u... <- discussion of commitment from Canada
https://financialpost.com/news/trump-says-tariffs-on-canada-... <- as of today (sorry for FT source but this is literally all across google you can just googla USA tarrifs)
0.2% of all US border Fentanyl seizures were on the Canadian border. That's almost literally nothing.
Also it's not only about drugs, but also humans smuggling (https://www.cbc.ca/news/canada/fentanyl-dr-smuggler-1.737348...), and overall border security.
- According to CIS, the number of Canadian crime groups producing synthetic drugs doubled between 2023 and 2024 - There's a lack of Canadian agents who are tasked at preventing this and current legislations make it very inefficient between federal and provincial law agents - There's an upward trend in Fentanyl seizures in Canada the last 2 years - Fentanyl is now being produced domestically in Canada
All of that is within the control of Canada with better policies.
Last year there were 45 lbs of fentanyl intercepted crossing into the US from Canada. Thats a backpack. There’s 500x as much coming from Mexico.
It’s unrealistic to expect that zero fentanyl will come into the US from Canada, and until that happens we will tariff all trade with them.
Tariffs are just a lever to get things done on the international front.
Canada has been neglecting security for a long time, so it's a wake-up call, and it's not a bad thing to put this out publicly to change things.
CBSA only inspects about 4% (some sources say even 1%) of all containers that ship in Vancouver. Quebec-Vermont* border control has been a joke for years.
* https://www.youtube.com/watch?v=Ug6Ryxza0tE * https://www.montrealgazette.com/news/provincial-news/article...
I mentioned in the previous post that Canada is willing to spend extra resources to avoid a trade war.
That means nothing to trump and this administration, you are trying to use logic to justify decisions made out of personal ego.
Edit: also wanted to add that more guns, drugs and people illegal flow into Canada from the US than visa versa.
Maybe the US should consider securing their own borders
Yes, and no.
Sure, Trump is using all kind of tricks to put pressure against his trade partners to secure some wins that will solidify him as a change agent to boost America, which will appeal to his electoral base. At the same time, it may actually bring good results to the US economy, bringing major investments in the country and negotiating better trade deals.
You may not agree with the means to get there, but you can't deny there's an argument to be made about his "America first" policies and why it could benefit the average people in the long run.
Of course the standard economic argument is that China using its GDP to make goods cheaper for our own citizens to purchase is better for us - they are subsidizing our economy. However it ignores the strategic disadvantage by our country losing its manufacturing capabilities.
The graphs may show economic advantage. It’s hard to quantify the long term strategic and militaristic disadvantage to not being able to make anything yourself if a world war occurs.
If things keep going the way they are going, that could describe the US just as well in a few short years.
"The company responded by intensively lobbying the U.S. government to intervene and mounting a misinformation campaign to portray the Guatemalan government as communist.[18] In 1954, the U.S. Central Intelligence Agency armed, funded, and trained a military force that deposed the democratically elected government of Guatemala and installed a pro-business military dictatorship.[19]"
This is not a one-time aberration.
Why do you think so?
This is overlooking the forest for one tree. The thing is, mean chinese manufacturing wages are $25k/year (purchasing parity adjusted! $15k unadjusted) for a 49h week.
That is the reason that so much manufacturing/industry has shifted there, not some nebulous "Chinese government subsidies" (not saying those are not a thing, just that they don't really matter all that much).
> It’s hard to quantify the long term strategic and militaristic disadvantage to not being able to make anything yourself if a world war occurs.
Certainly. But forcing low-skill industry to stay at a relevant size in a high-wage country is expensive business (compare agriculture, which is subsidized basically for exactly this reason) and not straightforward (see Jones act).
Presenting tariffs as a viable alternative to taxation is just beyond ridicule, but that has not stopped people so far either...
Most of Africa is just starting to slowly get there, Bangladesh is already very relevant for textile production.
I would expect the same basic trend to repeat that we saw with electronics manufacturing in 90s Japan: First cheap products move (very wage sensitive), then the local sector expands, wages rise with the whole local industry moving up the value chain, then at some point local wages become high enough for the whole process to repeat with the next low-wage country...
I think trying to block this trend off with tariffs is a futile waste of taxpayer money which american consumers are gonna pay for.
Spending tax money to keep some degree of self-sufficiency in critical industries (like with agriculture) can be a solid idea if done sparingly and cleverly, but that is not how the current US admin has approached this...
The bigger picture is that China invests in the development of an industrial chain. This has many aspects: infrastructure, education, training, housing, and of course tax incentives. The USA decided to stop investing in practically all of these. Even scientific research, the last area in which the US used to lead, is now in jeopardy from both sides: competition from China and internal cuts.
But I strongly disagree with your conclusion.
Lets assume that the US did all of those perfectly:
- Brilliantly educated workers with the perfect ratio of industry specific knowledge/experience
- Cheap housing near industry hubs built by the state
- The best and cheapest to use ports, roads and railway networks on the planet
- No tax on manufacturing workers income
Some of those are ludicrous/unrealistic for the US to provide.
But even if you managed to do all this-- that still does not make US manufacturing industry competitive. Because those US workers will still want a locally competitive wage instead of <10$/hour.
You are making some kind of logical jump here that I can not follow. I just listed basically the absolute best that the US could have ever done-- but even in that absolute dream scenario (tax free income for manufactoring workers? I mean in what world do you see something like that ever actually happening?), the US is still not competitive in a direct comparison, because US workers have just no reason to work manufacturing for 10€/hour (when they make ~30/hour right now).
You can stack all the incentives you want-- the gap in wages/standards is so large that apart from straight up paying the difference (in either tariffs or subsidies, and that is a lot of money), you are not going to make US manufacturing competitive in a head-on comparison.
I think you are reversing cause and effect here. Wages are not rising because things are expensive-- things instead become expensive because people are "rich" and can afford them.
I suspect (not an accusation!) that you would intrinsically like to see the US run healthcare and education in a government controlled way, at-cost, instead of allowing excessive private profits there (which I think is a good idea!).
But advocating for such changes in the name of making US manufacturing competitive is dishonest in my opinion, because I absolutely do not see those shrinking the wage-gulf sufficiently for US factories to compete head-on with China.
Furthermore, I don't even think you want to be competitive with China in this regard. Having a significant percentage of Americans working in/for factories to produce simple goods for 10$/hour strikes me as a step back, even if you would bundle this with a bunch of positive progressive improvements.
just think of china as another trader with more capital than you and pull yowrself up by your bootstraps.
Is the United States at risk of not being able to make anything ourselves? We have the second largest manufacturing output in the world.
https://en.wikipedia.org/wiki/Manufacturing#List_of_countrie...
I searched 'economics 101 strategic industries' and found this[1] within 30s which includes an overview of 'national self-sufficiency'. It presents the standard argument, including the parts you claim the standard argument ignores.
I personally favor decentralized planning over markets, but I find it unnecessary to slander economics.
--
1: https://www.adamsmith.org/economics-101
https://daringfireball.net/linked/2025/02/24/mission-accompl...
I don't pay enough attention to Apple's plans to judge if he is right.
* What type of jobs?
* Does the US have the required people, in terms of numbers and skills?
* Does this mean moving to US based fabs for the M-series chips?
* Is this actually profitable, or is this just a political move?
* Will this raise prices for customers outside the US for no justifiable reason?
* Does the US have the required people, in terms of numbers and skills? - "The company is opening up what it calls a manufacturing academy in Detroit, where it will help smaller companies with manufacturing. It already operates an academy for app developers in the city. It’s also doubling its manufacturing fund in the US to $10 billion." - Sounds like they are upskilling, and will count the employees of companies joining the academy as "jobs created"
* Does this mean moving to US based fabs for the M-series chips? - "[M-Series] chips themselves, however, continue to be produced in Taiwan.
* Is this actually profitable, or is this just a political move? - Define profitable. It is cheaper than paying tariffs.
Those 20,000 people won't be staffing the production lines. So how many manufacturing jobs, especially low skill, entry level with decent pay, will this create? The whole thing is framed in a way that makes it sound like Apple is creating thousands of manufacturing jobs.
doesn't matter because we have work visas
For 30 years, IT managers at blue chip US corporations have exploited the H1-B visa program by saying, "No," and then hiring a never-ending stream of barely-capable Java coders from programmer mills in India, take 5 times longer to make an app than it should have taken, get promoted, and leave everyone holding the bag with shitty web app that we all hate because it's too slow, too bloated, and doesn't work like it needs to. And the companies who can't get enough of that bullshit in-house just hire it out to sub-sub-contractors that do the same thing. Can we not invest in our native population and education systems this time around? I'm so tired of the fact that 90% of the IT staff in my Fortune 250 is Indian, and I know people who would be better at their jobs living in my home town. It hurts our community and our country, in the long run, and by the VERY same logic as re-homing our chip production.
People in the UK who are against immigration are often talking about Poles who moved to the UK after the EU and not Indian families who have lived in those neighborhoods for generations.
And they DO have families of their own here (and bring over their in-laws), and a lot of them don't integrate well, for a variety of reasons. At least a third of my neighborhood is Indian. They glare at me on the sidewalk when I wave. And most of them remain inured in their caste system, and are difficult and unpleasant to work with.
Again, all the same arguments about developing our own chips domestically -- which I doubt many people have a problem with -- apply to developing our own, better education pipeline to fully develop domestic software engineers.
This is not really a practical option. A big part of the M-series success is TSMC's lead in cutting edge process nodes. And Taiwan does not allow export of technology for the latest nodes. It is available only there.
I guess they could sell servers to customers who want to run the latest Apple Intelligence models on-prem, even though that probably wouldn't make much of a difference, since you probably still have to trust Apple.
Combined with a properly headless fork of their OS stack (think Darwin, not OS X Server) they could spin up a highly competitive solution using entirely "B-team" resources.
A modern version of the Xserve RAID for high speed flash storage could be very interesting.
The Mac Mini could be used as small blades.
or they could do something really wild, like take the Oxide rack scale approach and make something big for DCs.
But they might also want to get a piece of the prosumer homelab market that Ubiquiti is in?
That's along the lines of how it usually rolls whenever Apple tries to make something purely utilitarian, it's the most considerate and "fresh" look at a product, but ultimately designed to be used and then disposed when finished.
A purely utilitarian IT-appliance without a individual end-user doesn't seem to be possible in their product pipeline, you usually end up with something "Prosumer": Impressive on its own, yet of degraded maintainability and scalability.
It's like asking Bugatti to design a public transport bus. It would surely be an impressive bus, but not one you would want to maintain over years at a scale of hundreds.
> Apple said that it, together with Foxconn Technology Group, will later this year begin producing the servers that power the cloud component of Apple Intelligence — a system called Private Cloud Compute — in Houston. That marks a relocation, at least for some production, from overseas. Next year, it says a 250,000-square-foot facility for such manufacturing will open in the city.
[...]
> Apple will also expand data center capacity in Arizona, Oregon, Iowa, Nevada and North Carolina, all states with existing Apple capacity. The company confirmed that mass production of chips started at a Taiwan Semiconductor Manufacturing Co. facility in Arizona last month. Bloomberg News recently reported that plant is building chips for some Apple Watches and iPads.
Makes sense from a business perspective - there's significant growth potential for them as their presence in web tech is approximately nil.
Apple now offers access to Apple Intelligence Pro at 9.99$ a month.
LLM Siri cannot come fast enough.
For natural language processing you need a different kind of neural network don't you?
I don’t ask Siri for facts (just like I don’t ask LLM’s for facts). As long as it can correctly, understand what and when I ask to be reminded about something, that would be a huge improvement for me.
That and being able to map “Bedroom Fan”/“Bedroom Fan Light” to “Bedroom Fan Lights” without having to specify aliases (and even then it hearing me wrong).
I’ve see Home Assistant working with LLMs and it can understand groupings that I never explicitly defined which is very nice. I can say “Turn off all overhead lights” and it will find all my overhead lights and turn them off. Siri/Alexa can’t handle those tasks currently.
The precise mechanism LLMs use for reaching their probability distributions is why they are able to pass most undergraduate level exams, whereas the Markov chain projects I made 15-20 years ago were not.
Even as an intermediary, word2vec had to build a space in which the concept of "gender" exists such that "man" -> "woman" ~= "king" -> "queen".
Maybe I'm asking for an explanation :)
Since you seem to understand the mechanism, can you do a 3 line summary please?
Make a bunch of neural nets to recognise every concept, the same way you would make them to recognise numbers or letters in handwiting recognition. Glue them together with more neural nets. Put another on the end to turn concepts back into words.
For a less wrong but still introductory summary that still glosses over stuff, about 1.5 hours of 3blue1brown videos, #4-#8 in this playlist: https://youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_...
... Oh interesting. And those concepts are hand picked or generated automatically somehow?
> For a less wrong but still introductory summary that still glosses over stuff, about 1.5 hours of 3blue1brown videos
Sorry, my religion forbids me from watching talking heads. I'll have to live with your summary for now. Until I run into someone who condensed those 1.5 hours in text that takes at most 30 min to read...
Fully automated.
> Sorry, my religion forbids me from watching talking heads.
What about professional maths communicators who created their own open sourced python library for creating video content and doesn't even show their face on most videos?
You're unlikely to get a better time-quality trade-off on any maths topic than a 3blue1brown video.
He's the kind of presenter that others try to mimic because he's so good at what he does — you may recognise the visuals from elsewhere because of the library he created[0] in order to visualise the topics he was discussing.
[0] https://docs.manim.community/en/stable/faq/installation.html
There's also a playback speed slider in YouTube. I use it a lot.
A Markov chain (only using the probabilities of word orders from sentences in its training set) could never output a command that wasn't stitched together from existing ones (i.e. it would always output a valid command name, but if no one had requested a reminder for a date in 2026 before it was trained, it would never output that year). No amount of documents saying "2026 is the year after 2025" would make a Markov chain understand that fact, but LLMs are able to "understand" that.
Try asking ChatGPT to remember some obscure film you can only remember very hazy details about - really random stuff - I bet it will identify it for you I a few tries.
You can also totally miss spell words, use messed up grammar and it has no problems at all
Being able to chat casually with low latency, correct yourself, switch languages mid-sentence, incorporate context throughout a back-and-forth conversation etc. turns talking to these kinds of systems from a painful chore into something that can actually add value.
...that will grind your request to set email Vacation Mode through the world's worst speech-to-text, jam the text into Chat GPT, and spend the next three minutes reading you an uninterruptible 3 minute essay about violence.
I don't think it's been demonstrated that Apple could make Siri better with an LLM.
I had a birthday invite with clear date and time: so I asked it to add to my calendar.
It just said, “add what?” repeatedly until finally deciding it needed to send it to chatGPT to help. Which it did, then just returned the text in the image without taking any action. Then I say “can you add the event now?”
“Add what?”
So I try copying the text from the image in photos and giving that to Siri to add as an event. Surly this can work?!?
Nope.
It’s just utterly pitiful.
Even more annoyingly, sometimes there is a beep! ¯\_(ツ)_/¯
Apple is stuck and it’s AI will never be good enough until those creatives embrace it. Right now it’s disdain when mentioned.
> The problem is AI current best use case is creative work, art, music, programming
By “best” do you mean “marketable?”
Seems weird to see a bunch of creatives and creative professionals “disdain” a tool and still say it’s “best” for them…
So far it’s not though.
Then why doesn't it have a professional CAD application?
AutoCAD came to the Mac when Intel was shitting the bed (with aggressive OEM contracts for first party system integrators that prevented AMD adoption across HP/Dell/Lenovo-lines) and Windows 11 was being forced on users.
WINTEL played the monopoly game too hard and is starting to lose ground.
You love to see it.
But your statement isn't quite right. Fusion 360 runs fine in mac. I'm ex-Apple btw.
This is where it’s being pushed and marketed but I’m not actually sure it’s the best use case.
"Your 40 minute timer starts now".
It feels like it was a residual timer or something but I have never set anything like that - it is quite strange.
I’m serious, that’s what I use.
Somehow I don't think fealty will change its quality.
I.e. My preference for apple CarPlay supersedes my concerns on GPT running over my contents. Though the UI/UX has made it next to impossible to turn it back on.
What a world to live in.
How did you turn Siri off in the first place? That's where I'd start...
Settings -> Apple Intelligence & Siri:
Talk and Type to Siri -> turn all this back on
Allow Siri When Locked -> turn this back on
IMO, moving towards AI just leads to increased uncertainty and undesirable outcomes, which is something several journalists reviewing Apple Intelligence have attested to.
This is what our "AI accelerated" chips give us in return? What a disgrace
A command economy has different elements of the economy ordered around to do what leadership wants.
That seems a lot more likely.
We have maybe fifty or a hundred million people rotting away in areas where jobs are scarce and housing is plentiful, because we used government policy to shut them out of areas where jobs are plentiful and housing is scarce. We systematically exported jobs from places that aren't big cities because they can be performed overseas and our aristocracy can still profit from them by owning those people overseas.
I don't know if returning to a little more deliberate of an economy is even a partial salve for the place we've found ourselves, but I don't think this laissez faire thing is sustainable for a whole lot longer. We are overleveraged, and arrogantly delusional about our sway at the moment; "Ownership" is not some valuable skill. The fall of an anchor currency and global conversion to an alternate financial network would be a spectacular thing, an astroid striking terrain, which might leave craters on entire other continents from secondary ejecta. World wars have been fought over less.
I guess before is before Christopher Columbus
The desktop case for my 286, I could stand on it and it would not bend!
The Xserve, itself, was a 1U unit that was pretty much the same (or lighter than) any other 1U server (we also had HP and Dell servers that were heavy). The weight distribution could be weird.
That stupid drive bay was a proprietary nightmare. The disks cost a fortune.
However, from what I can see, this will be for "internal-use" servers. I don't think they will be selling iron; just services run on the iron.
Consider the cost of GPUs, losing what could be double digit percents on overhead might not make this very competitive. The macOS microkernel can still beat NT in some situations (like not having filters slowing filesystem down to a crawl), but it lags significantly behind the investment in Linux performance over the years by every other major company.
[0]: https://security.apple.com/blog/private-cloud-compute/
Doing a Mac Mini sized version for SOHO would be great, too.
One can dream, I guess...
You can't easily sell "Good / Better / Best" version of a single model, and tell "These are the options, take it or leave it". Servers are customized to the screws they come with and are expanded throughout over the years. So, the logistics are somewhat different for these kinds of devices.
Plus, macOS is not a CLI first operating system for server operations, and macOS Server is not updated for some years. Allowing Linux would be a different offering, and allowing macOS to work with all kinds of hardware from ordinary Ethernet to 100G+ Ethernet and 400gbps Infiniband (plus all the other interconnects) will be a fun exercise in testing flexibility of both macOS and Apple development teams.
So it's quite complicated. All servers are built to order SKUs. Dell keeps configurations "per server" in their databases, for example. If you have a Dell server, enter its service tag to support site, and you'll get the configuration of the device as it left the factory.
I mean places like Github and AWS are painfully racking up Mac Minis for their deployments and this theoretical server model would simplify everything. It also becomes an option for on premises AI inference using MLX, especially if they manage to get ANE support working in conjunction with the GPU for faster prefill.
The support and software stack for the server model would be the exact same as the consumer variant and they certainly wouldn't have special Linux offerings, Infiniband, and all that. If there's networking beyond their existing 10G it's going to be built into the board and they aren't going to support random 3rd party cards. The unit also doesn't need to be upgradable either.
The idea being there's no VMware, kernel or piece of hardware that can have backdoors built into unless someone files off the top of the chip and somehow probes the silicon
> Apple said that it, together with Foxconn Technology Group, will later this year begin producing the servers that power the cloud component of Apple Intelligence — a system called Private Cloud Compute — in Houston. That marks a relocation, at least for some production, from overseas. Next year, it says a 250,000-square-foot facility for such manufacturing will open in the city.
> The Private Cloud Compute servers use advanced M-series chips already found in the company’s Mac computers. Those chips themselves, however, continue to be produced in Taiwan.
Most countries don't have the resources to do much, but even then they can try their hardest not to be beholden to any single foreign country coughChina, Americacough.
That's the whole point of tariffs, to encourage domestic production.
Put another way, what is the difference between what you wrote and
>But US companies should be very wary of EU tariff blackmail that forces them to build production lines in the EU.
?
Whether a particular tariff is economically viable is a reasonable debate. Calling Trump's tariffs "blackmail" without assigning that epithet to all tariffs from whatever source is not.
This doesn’t stop Trump from pressuring and squeezing and making false claims to justify his standing point. Expect an unstable trade environment due to tariffs and retaliation. This will hurt small businesses that have even more complex environment to operate in.
VAT is what Americans call "sales tax" and US companies do pay and charge that.
I actually think Apple is a potential dark horse when it comes to AI hardware. What we have now is essentially a three-layered monopoly: ASML, TSMC, NVidia. This has been incredibly lucrative for NVidia. But, as we know, Apple doesn't like to rely on third-party hardware. They've invested heavily in ARM going back to buying PA Semi [1]. Apple replaced Intel chips (which originally replaced Power chips) with the M series in recent years. Apple is in the process of replacing Qualcomm modems in iPhones, which is not only a technical feat but a legal one given Qualcomm's patent dominance over 4G/5G.
Apple has the resolve for long-term initiatives that few other companies have. Apple Pay continues to chip away and get slowly better in a way that, if it were a Google product, would've been cancelled, rebranded, relaunched probably 3-4 times by now (Google Checkout, Google Wallet, Google Pay, Android Pay, etc).
Apple clearly sees AI as a strategic issue. They have loads of cash on hand to finance basically anything they want. And they won't want to be beholden to NVidia.
I expect Apple to have a significant impact here but it won't be tomorrow or even this year. It'll be over the next 5-10 years.
[1]: https://www.wired.com/2008/04/four-reasons-ap/
Then there's also the software issues. Nvidia has invested in GPU-based compute nonstop for the past 10+ years. Apple invested in Nvidia, then invested in OpenCL after abandoning Nvidia, then abandoned OpenCL for Metal compute which would eventually become the proprietary Accelerate framework. Nvidia's eggs are all organized in one, valuable basket. Apple's investments are spread out all over the place, with much of the time and money going into projects that don't even exist anymore.
Apple has the TSMC advantage, but that's just about it. Their GPU designs aren't comparably efficient or compute-oriented to what Nvidia ships today. Additionally, Nvidia will continue investing in places that Apple principally refuses to support. Unless a serious tide change occurs at Apple, they aren't going to get a fair competition with Nvidia.
Still in a AAPL long position since $170 and waiting for the re-introduction of Xserve with a new Apple Silicon-based OS. [0]
Probably "aiOS" (Just guessing).
[0] https://news.ycombinator.com/item?id=40278371
There's a lot of noise I can see behind the scenes on investor confidence. Noise as in "everything is fucked" sort of level of noise. Thus I expect this is being said to try and stop the AAPL stock collapsing in the upcoming recession that the analysts are predicting more than a tangible expansion and recruitment goal.
I also take issue with their being 20,000 people on the market who are still able to contribute something useful. They will be culled quickly and quietly down the line in the annual corporate lay offs.
It is not the time to make grand gestures unless you're trying to gain political favour, at which point any respect I have at least is gone.
Market opening today should be interesting...
The difference between the great leaders and the crap leaders is all in the details.
Even if the move forward with investment, they will be a bit of a 'late' mover, but will have had a chance to see what is working and what isn't working for everyone else.
1. JD Vance independent of Trump will have the same policies.
2. JD Vance will have enough popularity for a serious 2028 run. He might fall out favour with Trump as Trump tries to mount a bid for a third term, Trumpism might just generally lose popularity if policies lead to bad outcomes.
3. Dems don't figure their shit out. They should be able to take back some control in mid-terms, and then start to push their own policies, or at least credibly show that the most extreme policies from the exec branch don't have teeth anymore.
How do we know this isn't just crazy conspiracy theory? Because they already did attempt the same thing in 2020, and this time they had the chance to vet the VP candidate for this scenario.
Come back to reality.
Assume the human resources at $100,000 per head, and you get $2B/yr. Four years comes to just $8B for human resources. Assume land costs $10B. Assume construction costs $100B.
100+10+8 is $118B
Where is the money actually going?
Operational costs are going to be staggering. These are data centers that use as much power as entire cities.