Transitory to mid-late twenties then by 7pm/early evening it was low one-twenties. Someone said something on a call or the major holders got out for some other reason.
Until I can run frontier models locally on modest hardware, the hardware race will continue.
It’s kind of like when 3d graphics came out. We all kind of knew ‘the graphics are good, but not there yet’.
Eventually it does get to Crysis level graphics, and we all take a breather. But then it starts up again with Unreal 5 engine. We will get to photorealistic even if takes a 1000 years.
AGI is a very different level of problem to solve than ray tracing.
Photorealism is basically a straight shot, the algorithms for light are well understood, we just haven't had the hardware to do it in realtime, but the hardware was relatively easy to iterate on.
AGI is not the same. No, LLMs are not going to turn into an "Artificial General Intelligence" just because we keep improving them - LLMs inherently don't have the capability of thought, interest, reason, motivation, or anything we ascribe to human intelligence. And throwing faster hardware at the problem isn't going to solve that.
Sure, people are going to keep trying but I doubt anyone alive today will be seeing a true artificial general intelligence. In 1000 years? That's quite the goalpost.
We largely do understand how to replicate the brain though. Obviously not exactly but the brain is malleable so thats fine. The problem has been that we dont have the hardware to multiply matrices fast enough. LLMs are still orders of magnitude smaller than a human brain. Same problem ray tracing has.
Sorry but no, "multiply matrices" is not the fundamental problem blocking us from "Artificial General Intelligence", unless you believe in some alternate-facts description of AGI, as Sam Altman recently tried to foist upon the world by describing "AGI" as something capable of making money. That isn't the real definition of "AGI", that's a guy puffing up his company to try to get more money.
A truly intelligent machine would need the capability to reason, understand, and think, exhibiting cognition, sentience, and consciousness - not merely sift through training data for a plausible response to an input. That isn't intelligence, that's a fancy search engine.
LLMs and the entire premise behind them is incapable of producing AGI. They might seem like "AI", but they are nothing at all like AGI.
Nope. You could build an LLM using every transistor that ever existed of ever will exist, and it would still just be a fancy search engine with no cognition, no consciousness, no actual thought. People will have given up on the LLM hype long before actual AGI is ever close to being a thing. Do you even know how LLMs work?
No, we certainly do not - we only have a very basic understanding of the biological component, nowhere near enough to replicate consciousness. You're fooling yourself if you think it's just a matter of throwing hardware at the problem.
Meta seems like a zombie company, no. They have one product, WhatsApp, thats isnt monetized or at least isn't monetized where I am. Their other product Facebook I dont use nor do I know anyone my age who does, and Instagram, which they are ad-polluting to the point of un-usability.
I'm now one of those people, the ones talking about how they dropped FB.
Eventually I logged on and noticed that as I was scrolling, like 2/3rds of the content wasn't from my friends at all, not even re-shared stuff, it was all ads or shorts from randos or random group pages that somehow appeared on my feed. And of course a lot of the things from my friends were, in fact, re-shares.
I felt like, if only like 20% or less of the content is original stuff from my friends, what's the point of even being here? This is terrible. So I gave up.
When I look at Facebook now, I immediately click Feeds > Friends. It's not 100% perfect, but so so much better. And I see what I "need" with so much less wasted time.
I think not "zombie", but rather, "mature". Which is almost as bad from the point of view of people who bought their stock thinking it was still a growth stock.
Their user base can't get much bigger, they haven't ever really had a hit product after the first one, and they have a record of pouring billions into things (like AR) that never pay off.
Now, they have done ok with buying other companies and keeping them running (Instagram), which to be fair lots of tech companies have failed at. So I don't think they're a zombie. But they may be something more like IBM, Cisco, or Oracle; a mature tech company that will go up and down, but whose big growth days are behind it.
A few hours in the "metaverse" destroyed any faith or confidence I might have had in their ability to develop usable products that might interest people. It was like using a product that had been designed and developed by the wrong species with absolutely no interest in the actual user experience. Billions were flushed down a product toilet chasing something that just... stinks.
The experience is so overwhelmingly bizarre, frustrating, eery and depressing from a user pov that the only conclusion I could draw was that no one cared, that it was never actually tested, that no one actually was developing it for a user.
So I don't know. I don't think the company is functionally capable of developing usable, engaging products for people. If it were, it wouldn't have made that shambling abomination public, or it would have taken a very different form. I think it's just stuck enshitifying itself.
Frankly I don't really want to know more than I do about Meta, but it really does make me wonder whether there were actual voices ringing alarms from day one that the product was a dumpster-fire with horrific fundamentals, or if somehow the reporting on it came up all lilies from up-talking yes-yappers.
> they haven't ever really had a hit product after the first one
Threads is the fastest growing app in history.
And its insane growth showed quite starkly the raw power of Instagram and Facebook in being able to drive large volumes of traffic and bootstrap new products overnight.
I would be surprised if they didn’t try this again to launch a new dedicated TikTok competitor.
Fair. I am on threads and it’s fairly interesting.
However it suffers from what I’d call most of the content being unantheutic. No one is making content just for Threads. They’re sharing videos or tweets from other platforms and it’s their secondary platform.
I’m not sure there’s some niche this isn’t true. But to me threads feels like the personal version of LinkedIn. Lots of self advertising, very little authenticity.
a) We have an ageing population which is why you still see solid usage in Facebook. And their Groups/Marketplace features are strong drivers of growth.
b) Meta launched Threads a year ago which is at about 300m MAUs and every month taking more and more users away from X.
Zombie companies don't keep growing at astronomical rates. I've been hearing about the imminent decline of Facebook for the best part of a decade.... still waiting.
The market punished meta for Mark going rogue and ignoring wall st. advice during covid. The stock is up 7x since then. Clearly investors are happy to have mark even if we wastes 10 bil a year on reality labs.
Reality Labs lost another ~$15 billion over FY 2024, between that and the Vision Pro apparently falling short of Apples expectations it doesn't seem like VR is going anywhere fast.
VR has always sorta been a hedge against missing the next platform the way many companies missed out on the smart phone or social networking - it was never obvious it was going anywhere really.
All the AI research is under reality labs. Llama team is probably going through a few billion in capex alone. Nintendo doesnt produce most of their own games. Meta employees probably cost 5-10x as much on average compared to Nintendo employees. Honestly wouldnt surprise me if the average is above 10x
Ray Ban Meta is the first AR/VR product that's worth the price of admission imo.
The price isn't crazy, it's a very well put together product and I actually find myself using them. I wish I could use my AI of choice rather than Meta, but even so the AI presence in the sunglasses is actually helpful sometimes and the rest of the time is completely out of mind.
By far the best "AI assistant" hardware by a mile. Super impressed by the team behind it.
It's really hard for me to understand how Meta makes so much money, and why the stock is growing so fast. I wonder how analysts come up with such high valuation (or maybe they don't and it's just an AI bubble?). Meta isn't diversified and their products are clunky and buggy. How long will they be able to maintain their profit? are they capable to come up with something revolutionary enough to replace their profitable products?
It doesn’t sound like you understand what Meta is.
They are an advertising company that uses apps to drive traffic and collect data. And along with Google they dominate the space and are continuing to grow at a rapid pace. Whether their apps are clunky and buggy is irrelevant when their advertising platform isn’t.
And they will be able to maintain this profit for decades because most businesses are very reluctant to take risks with a profitable channel to market.
What people have to pay attention too is Ad Prices.
They are proxy for how much Attention is available in the world that Corporations/People are willing to pay to capture.
It has been going up for past 3 years and cannot go up forever given the competition in the online ad market with Amazon, Walmart, Netflix, OpenAI etc all rolling out their own Ad biz.
There is nothing predictable about what happens to Meta.
Whatever channel Just pay attention to ad prices. They have been going up on all channels.
Being a platform allows them to squeeze people. Only for so long . Before something gives. And the tipping point will be seen when Ad prices plateau or start dropping. Like home prices rhey cant go up forever because just like Land the pool of Attention is fixed.
To give an idea why, I’ll
present to you my own business case study.
I just bought $7,000 in ads at Meta for 2 months, for my retail clothing store that has 4 employees. We don’t even sell online. We are pure brick-and-mortar retail. I expect revenue to be 4x-10x that for that time those ads run. Whenever I buy ads on their platform, it’s always that kind of result. It’s better than my Google search ads because fashion isn’t something you necessarily seek to solve a problem. It’s something that someone else convinces you should buy.
Meta has billions of people staring all day on their platform, and they have ways for me to easily target the exact person that would buy our clothes, even if they aren’t searching for it.
I could do marketing in other ways, but this is the laziest, easiest way. And I’ve done things like direct mail and video ads and posters. I just have a full time engineering job and this is my side hustle. Fashion marketing is a lot of work, and Meta makes it easy.
And I believe there are millions of small businesses like mine that Meta is effectively taking advantage of. I know people in my exact fashion market of mom-and-pop clothing stores, some doing part time in their homes, some with small retail shops, doing the same thing of hitting “Boost” on their Instagram or Facebook posts. You don’t need to be LVMH to market fashion with Meta. It really does work at the micro-level.
Interesting. I see much less returns for a side hustle fashion brand. how much time do you invest in creating the ads? which platform do you use? Instagram or Facebook? or any others?
I use both. General Instagram marketing and for Facebook I create sale events and boost that.
There’s definitely a quality and established brand effect. If you have more content on your page or if you have better photos, your CPC are cheaper. I try not to spend more than a day a month on marketing.
Because its not possible mathematically. Those who pay more get the boost for the same target audience everyone else is trying to capture.
Notice it has nothing to do with quality of product or service. Just who pays the Platform more. If you are not getting attention. The platform tells you to pay more for it or that you be more engaging. This has nothing to do with product being sold.
Its classic middle man monopoly behavior. People cant do anything about it because its a system wide problem with a platform economy.
My own experience shows that the quality of ads, ad targeting, your website, your brand, your product, etc.. all affect advertising return-on-investment. I can see my ad return improving every time I improve the quality of one of the above.
You can spend a million dollars on advertising and get zero return if your product is absolute crap.
TikTok is getting banned but Facebook is way more harmful than TikTok. TikTok has never caused a genocide, but Facebook has, in Myanmar. Investigations revealed that Facebook's algorithms played a significant role in amplifying harmful content. Amnesty International reported that these algorithms "proactively amplified" anti-Rohingya messages, effectively turning the platform into a tool for hate speech and incitement to violence.
In October 2021, Frances Haugen, a former Facebook product manager, came forward as a whistleblower, and revealed that Facebook consistently prioritized profit over user safety and public good. She stated that internal documents showed the company was aware of its platforms being used to spread hate, misinformation, and violence but chose to hide this evidence to maintain user engagement and revenue. She asserted that Facebook's algorithms favored content that generated anger and division, which in turn increased user engagement.
The revelations forced Facebook to clean up its act — and lose revenue in the process — but now Zuckerberg is going back to inciting genocides for profit.
64 comments
[ 3.4 ms ] story [ 128 ms ] threadCapEx Full-Year 2024: $39.23B
Expected to increase to $60B-$65B in 2025, largely to support AI efforts.
NVDA bumped after hours on this news about 1.6% at the moment after a down day
It’s kind of like when 3d graphics came out. We all kind of knew ‘the graphics are good, but not there yet’.
Eventually it does get to Crysis level graphics, and we all take a breather. But then it starts up again with Unreal 5 engine. We will get to photorealistic even if takes a 1000 years.
Same goes for AGI.
Photorealism is basically a straight shot, the algorithms for light are well understood, we just haven't had the hardware to do it in realtime, but the hardware was relatively easy to iterate on.
AGI is not the same. No, LLMs are not going to turn into an "Artificial General Intelligence" just because we keep improving them - LLMs inherently don't have the capability of thought, interest, reason, motivation, or anything we ascribe to human intelligence. And throwing faster hardware at the problem isn't going to solve that.
Sure, people are going to keep trying but I doubt anyone alive today will be seeing a true artificial general intelligence. In 1000 years? That's quite the goalpost.
https://en.wikipedia.org/wiki/Artificial_general_intelligenc...
A truly intelligent machine would need the capability to reason, understand, and think, exhibiting cognition, sentience, and consciousness - not merely sift through training data for a plausible response to an input. That isn't intelligence, that's a fancy search engine.
LLMs and the entire premise behind them is incapable of producing AGI. They might seem like "AI", but they are nothing at all like AGI.
No, we certainly do not - we only have a very basic understanding of the biological component, nowhere near enough to replicate consciousness. You're fooling yourself if you think it's just a matter of throwing hardware at the problem.
Eventually I logged on and noticed that as I was scrolling, like 2/3rds of the content wasn't from my friends at all, not even re-shared stuff, it was all ads or shorts from randos or random group pages that somehow appeared on my feed. And of course a lot of the things from my friends were, in fact, re-shares.
I felt like, if only like 20% or less of the content is original stuff from my friends, what's the point of even being here? This is terrible. So I gave up.
I entirely agree with you.
Their user base can't get much bigger, they haven't ever really had a hit product after the first one, and they have a record of pouring billions into things (like AR) that never pay off.
Now, they have done ok with buying other companies and keeping them running (Instagram), which to be fair lots of tech companies have failed at. So I don't think they're a zombie. But they may be something more like IBM, Cisco, or Oracle; a mature tech company that will go up and down, but whose big growth days are behind it.
The stock did +50% in 6 months.
Just the facts
I would argue that Meta is just reaping the benefit and profits from some very old IP, some of which they don’t know how to monetise.
The experience is so overwhelmingly bizarre, frustrating, eery and depressing from a user pov that the only conclusion I could draw was that no one cared, that it was never actually tested, that no one actually was developing it for a user.
So I don't know. I don't think the company is functionally capable of developing usable, engaging products for people. If it were, it wouldn't have made that shambling abomination public, or it would have taken a very different form. I think it's just stuck enshitifying itself.
Frankly I don't really want to know more than I do about Meta, but it really does make me wonder whether there were actual voices ringing alarms from day one that the product was a dumpster-fire with horrific fundamentals, or if somehow the reporting on it came up all lilies from up-talking yes-yappers.
Threads is the fastest growing app in history.
And its insane growth showed quite starkly the raw power of Instagram and Facebook in being able to drive large volumes of traffic and bootstrap new products overnight.
I would be surprised if they didn’t try this again to launch a new dedicated TikTok competitor.
However it suffers from what I’d call most of the content being unantheutic. No one is making content just for Threads. They’re sharing videos or tweets from other platforms and it’s their secondary platform.
I’m not sure there’s some niche this isn’t true. But to me threads feels like the personal version of LinkedIn. Lots of self advertising, very little authenticity.
b) Meta launched Threads a year ago which is at about 300m MAUs and every month taking more and more users away from X.
https://www.tomshardware.com/pc-components/gpus/intel-ceo-sa...
Maybe Meta will get lucky. Its better for company to bet on something than do nothing for years like Intel and cry about other's "luck"
For reference, Nintendo revenues somewhere around $13-14b.
Meta is losing almost as much per year as Nintendo has in total assets.
Even with acquisitions I can't make the math make any sense here.
I refer back to this very bearish blog post from the acquisition of Oculus predicting that Facebook/Meta would be a poor steward of VR. https://assayviaessay.blogspot.com/2014/03/virtual-spaces-re...
The price isn't crazy, it's a very well put together product and I actually find myself using them. I wish I could use my AI of choice rather than Meta, but even so the AI presence in the sunglasses is actually helpful sometimes and the rest of the time is completely out of mind.
By far the best "AI assistant" hardware by a mile. Super impressed by the team behind it.
Not sure how the future will be, but I’d count on Meta being the largest company in the digital socialization space for decades.
And there's unlikely to be much pushback from the new administration if they want to buy any competitor who appears.
They are an advertising company that uses apps to drive traffic and collect data. And along with Google they dominate the space and are continuing to grow at a rapid pace. Whether their apps are clunky and buggy is irrelevant when their advertising platform isn’t.
And they will be able to maintain this profit for decades because most businesses are very reluctant to take risks with a profitable channel to market.
They are proxy for how much Attention is available in the world that Corporations/People are willing to pay to capture.
It has been going up for past 3 years and cannot go up forever given the competition in the online ad market with Amazon, Walmart, Netflix, OpenAI etc all rolling out their own Ad biz.
There is nothing predictable about what happens to Meta.
And so none of the channels you listed apply.
Also it's very easy to build an Ad biz. It's very hard to build one that is competitive.
Being a platform allows them to squeeze people. Only for so long . Before something gives. And the tipping point will be seen when Ad prices plateau or start dropping. Like home prices rhey cant go up forever because just like Land the pool of Attention is fixed.
Still relevant because ultimately ads show up on these app, and users may eventually migrate to other products.
Also relevant because it may shows their inability to write new products.
I just bought $7,000 in ads at Meta for 2 months, for my retail clothing store that has 4 employees. We don’t even sell online. We are pure brick-and-mortar retail. I expect revenue to be 4x-10x that for that time those ads run. Whenever I buy ads on their platform, it’s always that kind of result. It’s better than my Google search ads because fashion isn’t something you necessarily seek to solve a problem. It’s something that someone else convinces you should buy.
Meta has billions of people staring all day on their platform, and they have ways for me to easily target the exact person that would buy our clothes, even if they aren’t searching for it.
I could do marketing in other ways, but this is the laziest, easiest way. And I’ve done things like direct mail and video ads and posters. I just have a full time engineering job and this is my side hustle. Fashion marketing is a lot of work, and Meta makes it easy.
And I believe there are millions of small businesses like mine that Meta is effectively taking advantage of. I know people in my exact fashion market of mom-and-pop clothing stores, some doing part time in their homes, some with small retail shops, doing the same thing of hitting “Boost” on their Instagram or Facebook posts. You don’t need to be LVMH to market fashion with Meta. It really does work at the micro-level.
There’s definitely a quality and established brand effect. If you have more content on your page or if you have better photos, your CPC are cheaper. I try not to spend more than a day a month on marketing.
Notice it has nothing to do with quality of product or service. Just who pays the Platform more. If you are not getting attention. The platform tells you to pay more for it or that you be more engaging. This has nothing to do with product being sold.
Its classic middle man monopoly behavior. People cant do anything about it because its a system wide problem with a platform economy.
I think its more a question of letting small businesses know, that it can become a trap if they over rely on these Platforms for sales.
You can spend a million dollars on advertising and get zero return if your product is absolute crap.
In October 2021, Frances Haugen, a former Facebook product manager, came forward as a whistleblower, and revealed that Facebook consistently prioritized profit over user safety and public good. She stated that internal documents showed the company was aware of its platforms being used to spread hate, misinformation, and violence but chose to hide this evidence to maintain user engagement and revenue. She asserted that Facebook's algorithms favored content that generated anger and division, which in turn increased user engagement.
The revelations forced Facebook to clean up its act — and lose revenue in the process — but now Zuckerberg is going back to inciting genocides for profit.
Facebook is the app we should be banning.