So for some context, "flying on instruments" is considered much better in flight. A fair number of novice pilots kill themselves because they get lost in clouds and get emotional and think they aren't level anymore.
The instrument tells them they are level, but they get paranoid and either stall or crash because they don't trust that the machine is more reliable than their emotions, even when it's drilled into them.
Very interesting. This sounds like a response to the Mighty [0] launch which Paul Graham has been defending on twitter recently after an outcry from some of the 'hardcore' developers such as Jonathan Blow [1] and Casey [2].
Do you think low paid workers are going buy a cheap laptop for 500$ and pay 30$ or 50$ per month for Mighty instead of just getting a more powerful Window laptop for like 1000$ or 1500$?
M1 macs can be had for $999, and they can even be bought with cheap financing options to convert it to a monthly payment.
Maybe there’s a market for people stuck on old computers but whose companies still spring for super fast internet and $50/month SaaS bills per user instead of just spending that same money (or less) financing the laptops they actually need, but it would be a small market.
I think that corporations may want to force all their users to use Mighty so that they can control exactly what goes on in the browser -- it is easier to pay Mighty to "virtualize" the browser, than it is to keep all computers up to date and without malicious extensions.
Maybe even prevent browsers from downloading files onto the local computer. A full recording of each user's sessions. Integrated password manager that uses the Mighty login to tie them all together.
So I view it as valuable to corporations and thus mighty falls into the B2B category of company which makes software end users hate, but corporations love.
> easier to pay Mighty to "virtualize" the browser, than it is to keep all computers up to date and without malicious extensions.
Not denying your other points but Amazon Workspaces[1] is a product that perfectly fits their needs.
From a top-level exec's stand point I would imagine they would be more willing to buy something like Amazon Workspaces which gives them 100% control and peace of mind Vs piecemeal approach such as browser, conference call client etc.,
Graham does this often. Instead of directly saying what he wants to say, he beats around the bush. When Allred was getting criticized for shady business practices, Graham wrote haters and thanked him and Musk and a bunch of other people.
Not sure why he does it, it is very underhanded, as though he's afraid of actually getting in a legit argument with the person. Easier to pontificate form his essays, have them posted to his website, then have the mods defend him.
"Essay is french for attempt." There are good and there are bad attempts, of course.
I think he genuinely believes these issues are much broader than a single company. So he wants to encourage discussion about the general issue, instead of sparking even more unproductive, repetitive argument about a single company which would overshadow the broader and more interesting issue he sees.
(As we can see here, where a quarter of the comments are debating the business model of Mighty.)
If a “Crazy New Idea” (CNI) is getting VC funding from establishment tech capital to deliver a twist in an existing product, it doesn’t seem like it could be so crazy.
Reading this I thought of CNIs like the Internet of the 1980s, women’s suffrage in the 19th century, Project Mercury, The Eiffel Tower.
The CNIs gaining ground today in tech seem to be crypto, brain-computer-interfaces, quantum computing, and CRISPR gene editing.
If there was a similar offering as mighty for certain development tools including Xcode and android studio together, that would be awesome and something I would look into. Basically something which would reduce the build times.
From what I understand of both the recent tweets and both of their fairly radical/extreme (and mostly correct) views about abstraction being a net negative, the criticism is less "Mighty doesn't work" and more "We shouldn't need to use an app like Mighty to get good performance on a basic website".
Sure, but the disconnect seems to be that Mighty is advertised for people who essentially use their browser as an OS (I see Figma mentioned a lot), rather than people who just want to do basic browsing.
I think Mighty has unfairly taken the brunt of a lot of growing discontent with bloat on the web. I'm not ready to fork over $30+ for faster browsing, but I'm glad someone is working in this space, because I genuinely would like to rent a fast virtual computer for the occasional video editing task and think Mighty is a step in that direction.
At minimum it seems like I’d have to choose an instance size, an OS that was compatible with my software, install the editing software, mount a common storage location with my footage, etc.
What I’d like to see is a service that let me add my credit card, click a button, and launch a video editor on a high-RAM machine that I pay for the hour (including the software license fee).
My hang up with Mighty isn’t about their idea or their technology or their execution. It’s impressive to see what they’ve done.
My issue is that it’s not a product I could feel good recommending to anyone, at least at the high pricing that was proposed in the last discussion. In an era of $999 M1 Macs (and even cheaper AMD laptops) and readily available financing options, it doesn’t make sense for anyone to throw their money away at a SaaS service that simply cannot perform as well as local Chrome on a modern machine.
I could see the narrow use case for limited situations where someone has
1. Weird IT department restrictions that require them to use old, slow computers but also
2. Budget rules that allow them to spend monthly money on a SaaS but not on financing the hardware they need to get their job done and
3. Guaranteed high speed internet all of the time and
4. An IT/corporate security department that is okay with them sending all of their keystrokes, login info, and browser data to a 3rd party service
Surely this situation exists, but it still feels like Mighty is targeting a broader audience by providing untrue claims about remote thin client technology somehow being faster than a halfway decent local machine. I’d feel equally uncomfortable if Stadia was charging $50/month while claiming to be lower latency than local gaming.
The counter arguments about disrespecting hard working startup founders or doubting visionaries feel like a strawman response to legitimate questions about the value of their service. The technology and execution look to be good, AFAICT. It’s the product, pricing, messaging, and value that I can’t recommend.
I was going to try Mighty but even in my somewhat relaxed work environment item 4 is not really solved by Mighty apart from "we will not access your data". Obviously they are very smart and will surely be working on E2E encryption but they do not seem to have it.
The whole value proposition is that the beefy remote machine is handling the rendering, isn't it? At least half the performance of a web page is wasted on DOM re-shuffling.
Not to mention, it's not like Chrome can run Javascript without looking at the data, and if they're not running Chrome, they'll start getting into compatibility issues.
If Mighty behaves like a full replacement for a modern web browser on my system, it will have to give the web pages I visit (opt-in) access to my GPS, camera, microphone, Bluetooth, battery information, and other peripherals, I sliding all of my mouse and keyboard input. It will have to allow me to upload or download files from the web page. Also, all my tabs are running on the same remote VM, so this only protects other systems, not my banking or JIRA.
Sure, it's the ultimate sandbox for the code itself, it probably protects me perfectly from Meltdown attacks against other programs on my main system, but I don't think that's enough to call it 'endpoint security'.
I don't have the faintest idea, but the notion of having my session cookies and credentials stored in a VM somewhere makes me very nervous and I am not even half paranoid.
So they need to figure it out otherwise it's only valid for personal use, which is not exactly the money is, imho.
Maybe some self hosted solution for corporate departments? Or waiting for an acquision from Citrix or VMWare...
No but it makes a lot of sense for Mighty App to be the middleman between your computer and the Internet. If I was them I’d just give it away for free and sell your data to Facebook (anonymously of course!)...
I agree with your stance. I was confused to come here and read that people thought he was referring to Mighty. I think Mighty is a pretty small, easily proven/disproven, not-crazy idea, not-that-new idea, with niche applications. To me that is a bit different than how this article reads. I think the debate is market size. The only real debate I hear with Mighty is whether it is good for <10M people worldwide or >1B people worldwide.
>4. An IT/corporate security department that is okay with them sending all of their keystrokes, login info, and browser data to a 3rd party service
Or just an IT/security department that is blissfully unaware of the shenanigans that some of its users are up to. Especially if people are accessing this from outside of the corporate network to start with.
My company uses webmarshal which fronts requests and blocks pages deemed not good for productivity/data security. In that type of situation you already have an intermediary so mighty could make sense - make everyone use mighty and give it a custom blocklist. Obviously not a situation that employees would be a fan of, but it’s something that corporate is already doing, and mighty could improve the experience.
There was this product[1] to browse internet offline by downloading something called “Web Packs”. This was back in 2005 just when I was graduating when I spoke to the founding team. They were naturally quite confident about the product taking off. Something seemed off to me I couldn’t point out what so I didn’t take the job offer. After all these years I realized the source my discomfort. They were actually betting against the speed of internet getting better. While most of the businesses like Amazon, Google were betting for the internet tech to improve this product did exactly the opposite.
To me Mighty sounds like a similar category of product. They are betting that PC/Laptop/Mobile hardware will stagnate from this point on. Exactly when Apple has launched M1, which blows the previous version out of the water, at a non crazy price. From this point it’s a matter of time other hardware also catches up in terms of performance and price.
Besides, yet another company to handover my entire browsing history and data for purported improvement in latency? I don’t know.
"They are betting that PC/Laptop/Mobile hardware will stagnate from this point on."
Me playing the devils advocate...
Perhaps not //forever//, but perhaps for the next few years (enough for them to make some good money).
Why? Chip shortage, GPU shortage.
Anecdotally, I just bought a computer for $7500, that, two years ago would have cost about $2500 (granted, it is the new tech, but even so, new tech of this tier two years ago would have been approx. $2500). On top of that, have to wait two months for it to be assembled. (ouch)
I agree with you that speed will probably not stop increasing, but if prices continue to go up, or even just level off, few people will be able to afford the new tech.
That being said, would I invest in Mighty? No since I agree with you in general....and agree with you about the zero privacy issue...very regressive position for someone like PG imo.
That assumes that AWS/Azure/GCP won't have the same issues getting chips to run Mighty on that PC manufacturers have getting chips to sell to consumers though. If the shortage keeps up, that is not necessarily true.
What if the future is that you don't necessarily just buy a machine in the future.
What if, when you purchase your machine, it includes a host of features including a could suite of things like; VPN, Cloud Browser, An actual amount of decent cloud storage, portable applications, hosted in your cloud bucket, but accessible from any device (If I have Illustrator, I can just run it from any machine - not just my own. I can allow guest access to my paid licenses - like Say i want my brother to be able to draw stuff on my Illustrator license while I am not using it.
Etc.
Imagine if instead of that $7,500 "computer" you bought a $7,500 'stack' and the access terminal you happen to be using most is the physical laptop that youre used to.
I tried to build something similar to this more than a decade ago (2007 or so - but wrote about it in 2004).
My biggest issue is better information/knowledge management.
I have TBs of files and data strewn all over. We should be focusing on "my information" and let me manage and secure that - and access it from any device easily, and securely.
If augmented reality tech works better the necessity of cloud as your machine will become ubiquitous.
However I think we gonna do have RTX in our mobile devices - AR glasses, whatever modern mobile tech - because of decentralization (security/ethics purpose)
> They are betting that PC/Laptop/Mobile hardware will stagnate from this point on
No, they are betting that no matter the processing power of computing devices of the day, developers will always push things to the limit and ship products that can end up being slow on devices used by a large share of the population.
This has always happened, because software product makers always want their products to have as many features/capabilities as possible, and release them to market as quickly/cheaply as possible, which means there are lesser incentives to limit functionality or to invest in performance optimisation.
There's no reason to believe this trend will cease, so Mighty is offering a service that allows people to get much higher performance of their webapps without always needing to have the highest-power computing device in their possession.
Agree, it's an arms race at this point between end consumers and app developers. However, it also means once Mighty becomes popular they have to contend with developers overwhelming their computing resources as well.
As long as Mighty keeps providing a better experience than the local browser on the average user's own device, they'll have utility for a significant number of people. Developers still need to provide satisfactory-enough performance to the bulk of their audience. Mighty is for the users in the right tail who are willing to pay for better-than-satisfactory. There's no reason to believe that right tail will ever disappear.
> developers can now ship any horribly optimized thing
No, they can't. Only the right-tail power users will ever pay for this. Unless almost all of a product's users are in that group, the developers still need to offer satisfactory performance to the rest of their users, or be vulnerable to a new competitor.
> Centralized processing of all computing, would just make current issues like censorship, security issues worse.
People keep saying this. I doubt even Mighty believes they or their paradigm will take over "all computing" - just the niche that really needs it. Even a small market share can make them a big/successful company, but without making the whole internet significantly more centralised than it already is. I don't get the panic.
Couldn't they instead be betting that despite improvements in hardware speed the web will bloat faster than the hardware improves? Betting on that doesn't sound that crazy to me.
That is indeed one of their premise. However that also has, perhaps unforeseen, implications. To start with, their VMs have to be at least one step better than the end consumer's machine. Can they do it without increasing their price? Secondly, there's no guarantee that developers will overwhelm even Mighty's cloud resources?
All that said, it's worth noting that the first set of customers seem super impressed. So maybe Mighty are on to something.
The Mighty model is timesharing CPUs. At least 80% of the time my computer isn't even used. And when I use it, I doubt average load is over 25%. So the Mighty model has a 20x utilization advantage by those numbers.
That should easily support twice the peak CPU power for me when I need it.
Another thing is that the Mighty CPUs will (presumably) be upgraded continuously, while my laptop CPU gets no faster after purchase. If that makes me only buy a new laptop every 4 years instead of 2, I've saved a lot of money and hassle.
I'm not saying this means Mighty will conquer the world. But there are reasonable arguments behind the model. Especially if you assume bandwidth will keep improving.
And I've responded with the reality of the business model. There's no "CPU sharing" on "my computer". It's a beefy VM in the cloud.
Well, it most likely shares the CPU with other VMs, but that depends on the cloud, and the instance. And since the browser is always open, Mighty will always run that VM, with no sharing: https://news.ycombinator.com/item?id=27063554
Hmm, I feel like this is missing that those cpu workloads are "bursty", and probably all burst around the same time for a given region. This analysis of unused computer time assumes they can sell your unused time to someone else, which either means network latency to another region or more capacity in the same region as you when you have similar usage times as everyone else. I have no idea if this idea works but I don't think it does for that reason.
Yeah, but I think it's better to think about it like virtual/remote desktops, where the granularity is a browser tab.
At least, that's the only way it makes sense to me. I might understand that to do 4k high res 3d on my phone - my phone will need help. And once that "help" is available - my TV, and my tablet and my smart watch can make use of it.
But I'm not generally willing to trade latency for cpu time sharing. What is interesting is an always on, always working desktop session. It's why I like screen/tmux and ssh, rdp - and would consider running a Linux terminal server, so my laptop(s) and desktop(s) could be a disk less, stateless thin client.
Make the observation that the browser tab is the new process/application granuality - and it makes sense to host tabs in the cloud.
Personally I'd want to self host it - but the idea doesn't sound quite so inane.
We have game streaming services from NVIDIA and amazon has had Amazon workspaces. With the global chip shortage which will affect a large amount of people I think that this would be the best time to prove your product to have a good effect.
Not only that but a great amount of people here live in a literal tech bubble and we believe that everyone has a reasonably fast laptop or can setup a large workstation.
> Not only that but a great amount of people here live in a literal tech bubble and we believe that everyone has a reasonably fast laptop or can setup a large workstation.
By that same token the average persons browsing habits are also not the same. Optimising for Chrome power users who have eleventy billion tabs open and run hefty web apps isn’t the normal usecase. My families browsing is mostly taken care of on a 2012 MBP and a couple of ancient iPads. Mighty would be functionally useless for us.
Heck my entire day is spent in Chrome working in Drive and writing code in our own development environment in our web app. I am a browser power user in that sense on a three year old gaming laptop and I don’t understand the Mighty usecase. It’s a niche within a niche at the moment although I understand the appeal of the business model.
The business model is obvious. Data. The new petrol.
The use case is the question. The general trend against personal computing. The naïveté of todays startups and VC in general.
But this is nothing new. It will pass, some people will get rich, some will lose money.
Do I like it? No. Do I care? No. Why? Because I have learned my life lessons well. If something survives as tech standard, it will be used. So I focus my attention on not being early adopter on any kind of tech.
Geforce now or even renting an entire computer with shadow is for graphics performance, and is not 30 bucks a month. You also won't be entering any sensitive data while playing games.
I think it's a cool thing they've done, but it requires you to put full trust into them as a company.
It is fascinating to me how quickly people reject the premise of Mighty, even after PG lists all the good reasons for replacing judgment with curiosity.
I’ll admit, my initial reaction to Mighty was “I can’t imagine it ever being faster than my behemoth PC.” But then I stopped for a second and got curious. What is Mighty, really? It is a thin-client. That’s it. And are thin-clients a bad thing? Well, if latency is an issue, sure. But “high-ping” is a somewhat solved issue, whether in multiplayer gaming, in terminal utilities like Mosh, or even in optimistic GraphQL mutation updates. Where’s the use-case where near-zero latency is vital? The only cases I can think of are games like Rocket League: fast-twitch games where even the latency between my controller and my PC is something I happily spend hours optimizing — where latency prevents the necessary feedback loop for learning (akin to trying to learn how to hit a baseball while drunk).
But beyond near-zero latency use-cases, why would a thick client ever be better than a thin client? At the edge of performance, this question is easily answered: I would never attempt to train a PyTorch model on my admittedly powerful GPU. That’s what the cloud is for. So when it comes to my browser, why am I content to eat up memory and cpu-time with hundreds of tabs open that almost always include one or two that are broken, soaking up my resources, and have to be hunted down and killed off so that IntelliJ can return to its normal lightning-fast speed?
Might goes even further and asks why I would want to run IntelliJ on my machine at all. Wouldn’t I rather run IntelliJ like I used to run Vim over Mosh, where I never have to worry about storage space, about download/upload bandwidth, or about my computer becoming sluggish?
And that’s the killer idea here: that thin-clients almost always beat thick-clients. One could even argue that the entire internet is premised upon this reality.
I’d happily pay Mighty to try it out for a bit. Even if it doesn’t work, I’ve dropped more money of less fascinating ideas. At the very least, I’m rooting for their success, because it would change a lot more than how you consume content over the internet.
I think the key difference between a pytorch model and your browser is that you’re actively using and manipulating your browser and a pytorch model is a long running process without the need for second by second interaction.
And as you note the need for more power goes beyond a browser, IDEs, gaming, rendering, Bitcoin mining. Why just do it for the browser? You can, and some people do, remote into a more powerful machine for all of their work. This was the norm when terminals were true terminals. We could go back to this, but having your own computer historically had much larger benefits for people.
> why would a thick client ever be better than a thin client?
Because thick client distribute the load among many computer. With a thin client all that load is way more centralised.
> I would never attempt to train a PyTorch
No one is asking you to do that. We're talking about Mighty, a thin client that:
- runs a beefy VM in the cloud
- runs a single app, Chrome, in that VM
- streams video to the client
If a million people run Chrome on their laptops and keep it open for the entire day (and your browser is usually open throughout the day), that's... just a million people with their laptops.
If a million people run Chrome through Mighty, Mighty needs a million VMs always open, and a million video streams, also always open.
See how a thick client is better than a thin client?
The post is about crazy-new idea. This app is a RDP solution that is not a crazy or a new idea. It all depends on economic viability and marketing to the right people.
> It is fascinating to me how quickly people reject the premise of Mighty, even after PG lists all the good reasons for replacing judgment with curiosity.
I was curious about Mighty and looking forward to their technology. That’s not the problem.
The judgment came largely when they announced that it cost up to $600/year. It costs so much that it’s actually cheaper to buy a whole new computer if you might need it for a year or more. Once you put a price tag on something and ask people to pay for it, judgment is fair game.
This whole dismissal of people saying that they couldn’t justify the product for the price as some sort of anti-curiosity thing feels disingenuous.
If people will pay 2x or 10x the price of self-hosting for AWS, then I have no doubt that some people will pay 2x or 10x the price of a laptop for Mighty. (I wouldn't, but I also don't use AWS :) )
> It is fascinating to me how quickly people reject the premise of Mighty, even after PG lists all the good reasons for replacing judgment with curiosity.
His arguments are too vague to specifically defend Mighty. You can insert any technology and his arguments are neither valid or invalid.
> Where’s the use-case where near-zero latency is vital?
I would say typing is a pretty big one. It is extraordinarily unpleasant when typing lag is anywhere above maybe 50ms,and even worse when it is variable, like it would inevitably be if going over the Internet. It's even worse with mouse movements, where occasional spikes in lag can ruin your day.
This is a solved problem though, with optimistic updates. See Mosh, which does this for the terminal. If you've ever had to run Vim on a high-latency remote connection, it feels like having superpowers.
Sure, but it's not solved in Chrome, is it? Mighty is just a VM running Chrome as far as I understand.
Not to mention, the problem is fundamentally simpler in a terminal, with a very limited range of outputs. The problem of optimistically drawing the result of your input on the screen is much harder when that input could affect any portion of the screen in any way, like it can on a JS-powered web page.
> I’ll admit, my initial reaction to Mighty was “I can’t imagine it ever being faster than my behemoth PC.”
I haven't found Chrome slow on my mid-range PC (i5-7500). Or my phone (Pixel 3).
I feel like I must be in some parallel universe to everyone else talking about how slow Chrome is.
Just now, I loaded the first 6 links on Hacker News. 1 didn't load at all due to a server error. The other 5 all loaded in under a second (measured by DOMContentLoaded). I have uBlock Origin enabled (only in Firefox on the phone). Maybe that helps.
I can have 100+ tabs open without slow down if I want to. The bigger problem is I'm less productive with 100 tabs open because... there are 100 tabs open. It's just too cluttered.
I'm willing to admit Mighty might be a good product if people have this problem with slow browsers. I just never found this was an issue. Maybe it's a problem on low-end machines, but how many people have a low-end computer but are willing to pay $50/month for Mighty?
I just set up a new computer and I was wondering why Chrome and Firefox was so slow. Then I remembered I hadn't set up AdBlock yet.
I wonder if Mighty has ad blocking? It's an interesting question of which will be better for their business -- blocking or no blocking. (I won't try it because of the obvious privacy problems, because I use a fast computer, and avoid slow web sites. But it probably has an audience.)
>even after PG lists all the good reasons for replacing judgment with curiosity
PG wants you to think that they're mutually exclusive. They're not, but he has a product to hustle.
The problem (for PG) is that curiosity is not uncritical. Curiosity poses more questions than those in the category "how well does this product work?" Anyone actually curious is going to wonder about the problem space, not just one proposed solution.
Pretty obvious questions include: "how did software get us to the point we're exploring this as a design?" and "could the problem it seeks to solve be addressed in a way that eliminates assumptions about the solution space?", "What are peripheral ramifications of design decisions, and how much do I care?" "Would other approaches solve the same issues, have the same ramifications?" or "Are they synergies to leverage by trying multiple things in concert?"
I did not take Jonathan Blow's tweet to be saying it's a bad idea, or won't work, but rather that it's an indictment of the whole web stack that it's necessary.
Watch any of his talks and you'll know for him it's definitely the latter. He's complained about bloat and the web stack in particular plenty of times.
Also note he's currently developing a programming language called JAI that's trying (and seems like it might ultimately succeed) in being a faster and smoother alternative to C++ for developing games.
I have no dog in the fight either way, but I think it's weird that their demo product shots are on a Mac when part of their pitch is
> "50+ tabs without your computer coming to a crawl"
On Macs, people can just switch to Safari for free and solve that problem. Yes, Chrome is a memory hog. Stop using Chrome, don't send all your browsing data to a third party.
Perhaps Windows would be a better choice for Mighty demo shots, since there may not have a better option than Chrome for Windows.
I just bought an used Lenovo desktop that’s a few years old, but has an i7 and 32gb of ram. It handles a hundred tabs without blinking.
I think the bigger problem is the artificial constraint we’ve put on ram, why are we still selling computers with 8gb of ram?
I am actually in Mighty's market if it cost ~$10usd per month, not $50.
At our startup we have a small org of customer service reps. They basically live in four tools: Notion, Slack, Chrome, and Front.
These four tools have one thing in common: they are all Electron apps. They SUCK. On any windows computer slower than a $1.5K lenovo, our reps can't have more than ~10 tabs open before their computer starts to stutter.
One answer would be to buy everyone an M1 mac. However, most of these reps are not familiar with apple (they use PCs at home) + they are still a bit too expensive (we're in Mexico where they cost 1.5 - 2x).
I would love to be able to buy our customer service reps a setup (monitor + mouse + computer) for ~$500 USD and have them use Mighty to run our customer service suite for $10-$15 bucks a month. That would scale to a customer service org of hundreds.
At $50 bucks a month I should just buy them a mac.
IMHO Suhail's pitch about "running figma" is wrong. I'm happy to buy a designer any computer they want. But outfitting a CSR team of 50+ people? Mighty + a thin client would be amazing.
>At $50 bucks a month I should just buy them a mac.
Yeah, agreed. I kind of thought the savvy business plan would be to charge at a loss, make up the loss with VC money, gather a lot of market share (customers like you), then raise the price once the pain of moving off is too high
How long before the $500 Lenovos and Dells have same ballpark performance to the 2021 M1 Macs? AMD or NVidia will inevitably release an ARM chip that competes with M1, and it will just be 12-24 months before this is commonplace.
I don't buy the argument that performance will always get tapped out by developers. There is an upper limit to what even a horribly architected 2D web app needs to consume.
Maybe! Hopefully! But it begs the question: why is the situation so bad now? What will change? It's not like the apps I mentioned are inherently complicated.
I think the situation is bad because developers are expensive, and really expert devs that know how to optimize app performance are even more expensive. You mention Notion and Slack and other Electron apps. Not to disparage the teams on these projects as they are impressive from a pure UX perspective. But building these complex apps cross-platform native is particularly difficult today. So Electron is used as a shortcut. The usage of Electron allowed Slack and Notion to grow rapidly with a smaller team than they otherwise would were they to try to replicate in Windows, Linux and Mac as well as web... not to mention mobile. And once an app is established and has tons of features, as they both do, doing a rewrite is both expensive and risky. Big rewrites almost always fail, and they divide your precious dev resources even more.
You can probably buy them a VDI desktop for $20 per month (Amazon Workspaces and Windows Virtual Desktop on Azure are in this ballpark) and run Chrome on that. Especially with WVD, there are thin clients made specifically for it (or you can just use an RDP client on any OS).
They didn't reveal their true business model or pricing yet.
The $50 subscription is a way to simply not flood their servers while beta-testing them, and also getting a bit of money from rich/dumb early adopters.
Well, if the Lenovo or Apple M1 were 3.5 to 5 times cheaper, they'd also be a good deal, right? In fact, they'd be a better deal than a 5 times cheaper Mighty, I would wager.
I don't understand how Jonathan Blow of all people, figurehead behind several great video games that would have been impossible to run on a supercomputer 25 years ago, could possible have this take.
I'm not upset about Mighty itself - I'm sure it's a fine product made by some decent people.
No, I'm upset that Mighty is necessary. That we, as an industry, have failed so hard that multi-gigahertz mult-core machines with gigabytes of ram can no longer consistently and quickly render documents without offloading it to a central server.
I follow Blow's work and opinions quite closely. I am quite confident he was NOT criticizing Mighty specifically. Instead he was deploring the state of software engineering in general and web programming in particular. He is saying something like "I can't believe web engineering sucks so bad that a tool like Mighty actually makes sense". See his talk about preventing the end of civilization (!!):
Definitely highlights something I've always felt, which is that we're really making life complicated with all those tools and processes that are less than ideal.
I get the feeling business and the need to "ship" stuff is the thing really bringing technology down, even though it looks like it's moving it forward.
This is the first time I’m hearing of Mighty App and, at least at first glance, I wouldn’t categorize it as a “crazy idea,” in fact it seems pretty non-crazy to me - chrome but more efficient. It’s literally a faster horse. I guess the implementation is the crazy part...?
This is an interesting idea, but I feel it's not calibrated right -- I'm an academic, so I'd imagine I work with many "domain experts" (if you don't think so, that's of course another valid discussion). I don't think "if you bet on the entire set of implausible-sounding ideas proposed by reasonable domain experts, you'd end up net ahead.".
I hear a implausible sounding ideas all the time. I think one of the main points of academia is to give people the chance to explore those ideas. But they don't turn out good "on average", not even close.
Yes, because even when those implausible ideas fail, your more likely to learn something than when experimenting on an existing hypothesis with just a minor change.
Your framing of the question totally ignores opportunity cost. All that time spent working on ultimately failing crazy ideas is time you did not spend working on anything else, either marginal and likely or less risky/crazy.
Ending up "net ahead" does not necessarily mean maximizing the fraction of successful projects. I don't think pg is saying that implausible-sounding projects usually turn out good. He is saying that when they turn out good, they have outsized impact, precisely because they initially sounded implausible and therefore produce lots of new information if true.
The median implausible-sounding idea is bad, for sure.
But if you never bet on any implausible-sounding ideas, you exclude the chance of being an early participant in any big paradigm shift.
A good exercise is to back-test your rule against big ideas through their history. Would you have invested time in solving Schrodinger’s weird equation in 1925? 1927? 1940? Would you have invested time in public key cryptography in 1975? 1976? 1980? 1990?
People who got in early, got to make the big discoveries in those fields.
> A good exercise is to back-test your rule against big ideas through their history.
Everyone assumes they would always have been on the right side of history. It's basically impossible to know what you would have been like had you been born in those times. Better is to back-test your rule against paradigm shifts that have occurred in your own lifetime. What are some ideas you dismissed that turned out to work? What are some products or projects you have a revulsion against that are still succeeding years later? What caused you to reject them? What did you miss?
This is an excellent point, and I think this comment thread would have much less negativity if everyone asked themselves this question.
But then they’d have to own up to the nauseating idea of callously dismissing some truly spectacular sea changes, such as failing to buy 1000 Bitcoins for a few dollars in 2009. I think most people are too full of envy to do that.
How many can honestly say they kept an open mind to the handful of such paradigm shifts we’ve experienced in the last 20 years?
You only have to be right a few times in your life to make your career. Most people are right zero times, and hardly anyone is right 10 times. Von Neumann might have hit 10.
So regretting past mistakes is a big waste of time. Learn from them, but don't let it get to you emotionally. (All easy advice to give, and hard to follow.)
A very smart professor of mine once thought "They added pictures to hypertext, that's sort of cool I guess."
Looking backwards definitely biases you to seeing things you didn't see at the time. Even memories get tinged. But memories are probably the most accurate you can do.
I missed bitcoin entirely despite knowing many people who bought as I still don't understand what the appeal is. It doesn't seem efficient to me as a payments technology or from a privacy perspective. I kind of wish I bought early even though I probably would have cashed out way too soon, as the money would have been nice.
I liked Stripe due to my revulsion to PayPal but wasn't exactly going to be on them winning. I was pleasantly surprised. Ditto Shopify because of my dislike of Amazon stores. I think I underestimate the chances of good ideas and good products winning because I know it's very hard.
I took a few bets early in my career on start-ups. I think I underestimated the difficulty of selling into retail on one company that I still think had a good idea. The others ended up being acquisitions, although the products weren't continued. I still believe social notetaking is an interesting hard problem in need of a good solution but I don't think we had found it.
I think the big lesson for me is to not get taken in by the cult of personality around strong founders and focus more on the merits of the ideas. That can be tricky as you're not getting the full picture when you interview making it even harder to see a crazy idea as good for the right reasons but it's the right approach.
I've found that you can get surprisingly far in life simply by being willing to fix your mistakes and switch your bets once it becomes apparent you're wrong.
I initially dismissed the WWW as a disorganized mess when I first encountered it in 1994 - I liked Gopher better. By 1997 I had changed my tune and was all-in on learning HTML and CGI and even some of the new technologies like Java applets, RealAudio, and DHTML. (Note that Java and RealAudio were themselves eventually losers, but by then I was all-in on Javascript.) It felt like I had missed the boat in 1997; the dot com boom was in full swing and people were getting rich off much more sophisticated stuff all around me (TBF, I was in high school). But the WWW and Javascript/DHTML remained a lucrative career for 20 years afterwards.
A lot of people, once they take a position on something, dig in and don't change that position for a lifetime, then miss out on the opportunity of a lifetime because they dismissed it to begin with. But truly world-changing ideas usually have a lot of room to grow. If you're wrong for 2-3 years and then change your mind, you'll feel really stupid, like you totally missed the boat, but that's nothing compared to the 20-30 years left of growth that the idea might have.
But keep in mind each of those was one of n implausible sounding ideas at the time. Even if you might have been willing, would you have had the resources to put into it, or would they have already been wasted on implausible ideas we don't remember?
As you point out, it depends on how we define "domain expert". I may be a "blockchain domain expert" but I also need to understand what consumers need from a blockchain in order for it to be a profitable venture. My idea may be very innovative, but if it doesn't solve people's problems it's not going to make much of a difference.
In other words, a successful business requires not only producing an innovative product, it also requires knowing if people actually demand this innovative product. One half is the core idea, the other half is how this idea interfaces with the world.
In the realm of entrepreneurship, I think domain experts who also risk tons of their cash on crazy ideas should be looked at even closer.
Y Combinator and Tesla were seen as stupid and crazy for a long time. Meanwhile I’ve seen a lot of domain expert entrepreneurs do really crazy things that lead nowhere when it’s VC money instead of theirs.
There are too many factors that'll decide whether it'll be success or not. I believe timings and luck are often overlooked.
iPhone certainly not the first touch screen display, Microsoft did have one, I believe. However Apple's timing was perfect, there was increasing adaptation of wifi connectivity and 3g, making it able to browse anywhere. Of course smoother interface and beautiful design also contribute big.
AmongUs was lucky they got advertisement from streamer, at perfect timing that was (is) pandemic and lockdowns.
Tesla can be considered crazy because they try to emerge with unknown product, with unknown market and unknown possibility of success. If we look at the past, Apple's iPhone was crazy too, because it was new product, with unknown market.
The first iPhone famously only had 2G cellular internet. At the time 3G phones were not ubiquitous but were readily available (had one as a mid priced Nokia flip phone), so I’ve always thought this was a fascinating aspect of the iPhone story. I bet the decision was made to go forward despite the product being slightly hobbled precisely because of timing.
The iPhone was an iPod that could make phone calls in an era when phones didn't do much beyond making phone calls.
So it was a no-brainer for those already carrying an iPod and a phone, even if it didn't do anything beyond being an iPod that could receive calls and texts!
I think the key statement to realize here is this,
> Most implausible-sounding ideas are in fact bad and could be safely dismissed. But not when they're proposed by reasonable domain experts.
Off the top of my head, I vaguely remember that the latter half of the careers of many "greats" like Einstein, Fermi, Erdos were riddled with pursuing directions which bore no fruit (to this day).
I don't quite think that crazy bets work on "average". On average they fail. It mostly so happens that crazy bets are on "average" taken by people with a network that can attract able and passionate people, who can see through an idea to its conclusion. This is where experts come in.
Experts often rely on validation by their "community", and by the nature of community dynamics need to upsell. They naturally attract the kind of people needed to execute crazy ideas. But then attracting such talent also has a corrective effect on the originally crazy idea, such that it has a higher chance of succeeding by minimizing the blind spots, the unknown unknowns.
> Off the top of my head, I vaguely remember that the latter half of the careers of many "greats" like Einstein, Fermi, Erdos were riddled with pursuing directions which bore no fruit (to this day).
I don't know about Erdos, but the other had ambitious ideas, not crazy ones. Evidence that they weren't crazy is that the problems they were working on are still seen as the ones to solve by most of the physics community.
In fact, on the case of Einstein, he worked on much crazier ideas on the beginning of his career. I don't think any respected physicist ever thought about solving Brownian motion, and the photoelectric effect was an interesting curiosity but mostly of interest of "those inventors out there", not of researchers.
This is the whole "black swan" idea that Nassim Taleb writes about – events that are very rare and unpredictable but completely life-changing. This might be a poor paraphrase, but roughly: We're not very good at reasoning about these sorts of events, and domain experts are not immune. It's easy to believe you have more predictive or explanative power than you actually do, and to expose yourself to unacceptable risk as a result.
In this particular case, my argument is more along the lines of what this "team of experts" does is a conscious choice throughout.
Black swan events can probably describe a lot of things, but my understanding is that an individual or a team does not really set themselves up for black swan events. It just happens.
> I vaguely remember that the latter half of the careers of many "greats" like Einstein, Fermi, Erdos were riddled with pursuing directions which bore no fruit
"Such ideas are not guaranteed to work. But they don't have to be. They just have to be sufficiently good bets — to have sufficiently high expected value. And I think on average they do. I think if you bet on the entire set of implausible-sounding ideas proposed by reasonable domain experts, you'd end up net ahead."
Isn’t this pretty much what Thomas Kuhn said in his Structures of Scientific Revolutions back in the 60s? At one time that was hailed as one of the most important books of the 20th century.
"If someone I knew to be both a domain expert and a reasonable person proposed an idea that sounded preposterous" - my reaction would be to ask: that's unexpected to me, what is different?
I agree with the general gist of this, however the crux is that most ideas that turn out to be revolutionary just don't originate within the accepted circle of domain experts, exactly because being a domain expert makes you more unlikely to think out of the box.
Google didn't come out of Yahoo. Napster wasn't founded by music industry execs. Friendster didn't evolve into Facebook. Match.com didn't invent Tinder. It's often the curious outsiders that disrupt the status quo, not the well connected insiders.
Those are all excellent examples of the main reason why innovation can not be done in established businesses: the fear of cannibalizing the cash cow will stifle any real innovation.
I don't think so, at least not in any technical aspect. Pretty much everything about the tech was a logical extension of industry trends. It began life as a CD-ROM add-on for the SNES similar to Sega's SegaCD for the Genesis/MegaDrive. Outside the technology, the low license cost and relative ease of developing for it built a sizeable and varied library of games, which I think was a major contributor to its success, but neither of those things strike me as particularly revolutionary.
Bell Labs is a counterexample to what you said. It developed the first transistor, photovoltaic cell, fiberiptic, lasers, cctv and some other things. Oh, and cellphones of course.
Ditto Xerox Parc and modern user interfaces, ethernet and and laser printers.
Agreed, it wiuld be more accurate to say it is not possible in companies that are optimizing finances short-term. Bell invested a ton of money with the hope it would bring results long term. I have yet to see a company's board authorizing that now...
It’s also just as often someone who’s worked in an industry for a while and has a bunch of good ideas from understanding the problems. We put the college kids making big companies on a pedestal but most startups are from people who know the domain.
You won’t find a revolutionary import/export company that doesn’t have some knowledge about the logistics industry. But you are more likely to be paying attention to new technologies as they emerge that might allow you to come up with the revolutionary ideas. I’ve noticed as I’ve gotten older I’ve stopped spending my time fiddling with all my devices. As a result I’m less likely to realize when something new and exciting might be within reach, until it hits a higher point on the bell curve of maturity.
However, for everyone of those examples there are many counter-examples: Intel came out of Shockley, ex-IBMers went onto found several other companies (SAP, Peoplesoft,...)...
So Im skeptical of this outsider/insider characterization, IMHO it's almost always the individuals.
Yeah this feels like an example of a narrative that sounds compelling but could easily be overstated. How many breakthrough innovations have “insiders” made that aren’t as noteworthy because they are more expected? Google was given as an example by the parent comment. Google also made a lot of breakthrough innovations in the distributed systems and browsers space after their outsiders phase.
I agree. The iPhone is the perfect counter-example to this narrative: perhaps one of the most revolutionary pieces of hardware or software in the last 20 years, it came out of Apple, an established company at the time.
from my reading of the history of Yahoo they had plenty of good ideas that they destroyed because the problem was not domain experts but thinking they were in the wrong business (media!)
Why are all your examples software companies? Certainly not because most of the world's revolutionary ideas are from programmers. Perhaps there are benefits to domain knowledge for coming up with revolutionary ideas in, say, biology or chemistry or electrical engineering or literature or political science or manufacturing or medicine...
I'm curious to see a list of world-changing ideas that came out of domain expertise and those that the essay claims could be dismissed because the person was not an expert.
AirBnB, which is perhaps pg's favorite company, would never have gotten into Y Combinator without Michael Seibel and the endorsement of the JTV crew. AirBnB simply would not exist today.
So I hope people don't read this essay and think, "I'm not a domain expert, so I'll give up on this idea." Somewhat of a dangerous essay for that reason.
I don't think pg intended that crazy new ideas require someone to be a domain expert, just some expertise. Slight wording difference.
In the case of music, one could have domain expertise on audio files, compression and domain expertise about file sharing and domain expertise about younger people's listening habits. That's just having some insights into the domain, several as is the case with Napster. Music industry execs in those days probably had domain expertise on contract law, revenue sharing models, marketing and didn't have a clue about file sharing or audio compression.
Also want to mention the examples you mention seem to have not much "newness" to them to be "crazy new" in my opinion.
When I think of "crazy new" ideas, I think of things like stuff that came out of PARC in the 60's. If you go watch the Stanford 2 part series "How To Invent The Future" with Alan Kay on youtube [0] it goes into a lot of detail about them.
The challenge there with Xerox was they couldn't capitalize on every single idea that came out, just not logistically possible for a company to really do that.
> the crux is that most ideas that turn out to be revolutionary just don't originate within the accepted circle of domain experts, exactly because being a domain expert makes you more unlikely to think out of the box.
I don't know about that... look at physics for example. You have plenty of experts who are total mavericks and then a bunch of outsiders spewing crankery and messing up on the basics.
The thing is, most crazy new ideas are just that: crazy. For every great new invention there is the other 99.999% of bunk or snake oil.
It's very difficult to differentiate between an innovative idea and claptrap. Genius is difficult to recognise because it takes a genius to develop the idea in the first place.
The issue is that domain experts often have crazy ideas outside their domain. Typical example is a technical expert or a scientist having a crazy business idea.
Something I’ve noticed a lot lately is that the people who have crazy new ideas are often the absolute last people that should be communicating them to the world. I’m not sure if it comes from the insularity of academia or just a basic inability to write clearly.
My theory is that many domain experts haven’t needed to communicate with a lay audience in decades (if ever) and thus aren’t aware of their own baseline assumptions. Seems like a startup idea, maybe? Convert academic papers into comprehensible English. Two Minute Papers does this but it’s only for technology.
This probably applies to early Apple. Wozniak, while clearly a technical genius, needed Jobs’ communication and design skills to sell “personal computing” and make Apple a mass-market company.
I think many (but not all) domain experts end up that way because they are so interested in a topic, they've internalized all the jargon and assumptions required to get there. It takes quite a bit of effort (and practice!) to keep the viewpoint of the common person in mind as well.
Feynman loved explaining things, so he had to keep trying to explain them in a manner that the ordinary folk could understand, but he also loved physics, which kept him diving deeper, and playing with it.
Having charisma and knowing what people want, and how to sell was Steve Jobs portion of the game, once the initial technical hurdles were solved. Woz likes to minimize circuits and do clever things, that was his part of the game.
Here on HN, the balance seems to be keeping it focused enough on hacking technology, while still appealing to those who want to make money off of it in yacht loads.
It's all about balance between at least 2 domains.
Graham's obviously inspired by some Kuhnian paradigms here, but in Grahamiam fashion, he doesn't cite anything to back it up.
What I find interesting where he calls most people conservative is how when I'm discussing VCs, YC, etc with non SV people is we talk about how conservative they are. YC's business model for years was predicated on finding talent (kids) that was undervalued, underpaying them (10k was the initial payouts?), getting them to pack up and move to super-expensive monocultural SV like everyone else, and then when they made bukku money, lose interest in improving the service (Reddit, Dropbox). There's nothing original there, it's the business model of carpetbaggers and robber barons. How boring.
I often feel like Jonah Hill in Moneyball, a pariah for pointing out how ancient VC thinking is. Or maybe I just imagine it. Well, it is my experience that true deep domain knowledge can only come from years of insights. People without years of experience will be lacking maturity and/or won't have time to even consider those insights. Usually people who mainly care about money, influence "becoming powerful" (Graham's words about his protege, not mine) will jump at whatever shortcut they can take instead of spending the card work necessary for learning these deep insights PG is interested in.
With all due respect, Coinbase didn't require deep insight. It was the equivalent of "You like money, too?"
.
PG, you should stop the contrarian persecuted intellectual look. You're not the little guy anymore, haven't been for 2 decades.
Capitalism for all its flaws at least taps into what motivates people to work really hard and encourages it. It mimics nature itself. Some ideas might be mundane but others can be extraordinary, it’s par for the course.
As an example, Capitalism gave us a vaccine to a novel corona-virus within a year. These companies didn’t do it out of good will, they did it because they knew they could make money. Many socialist countries are struggling because it turns out it’s hard to force people to produce innovation without the right incentive structures in place.
If you want to blame anyone blame God for making human nature this way. Capitalism is just the system that is most effective at tapping human psychology to push people to produce the best possible work.
As of November last year (let alone anything since re: distribution), US governmental agencies paid $2.5 billion to Moderna to develop the vaccine and buy doses (and there already had been work done by Moderna on mRNA vaccines).
As of July last year (again, let alone anything since), US governmental agencies had paid Pfizer $1.95 billion.
So, basically, the government spent billions for something they then distributed 'for free' to the taxpayers.
That's an interesting example to pick for 'capitalism'.
The companies you just mentioned all got Paid at whatever market rate they/the market set.
Just because it was the government that paid them, why isn’t that capitalism?
Socialism and Communism is when you tell people what they’ll charge for the greater good of the people. At least that’s what it is to me, someone from a former communist country.
Edit: not only did those companies get paid to invent it, but others got paid to make it, and yet others got paid to distribute it (pharmacies etc).
Really? I would have expected something a bit higher than $20-$25 a dose, given the limited supply (it was pure research at the time), and COVID was decimating the country (so extremely high demand). I mean, a single dose of insulin, requiring no research, and no protective IP, and costing cents to produce, will run you or an insurance company hundreds of dollars. Given the huge demand, and the non-existent supply at that time, there is -nothing- in the market that would have capped the price for the vaccine at $20-25, except either human good will (which is not a market factor; again, see insulin), or the very real threat of government action.
>> At least that’s what it is to me, someone from a former communist country
And voting Republican is what Democracy means to plenty of people living in the US. If we actually go to what the dictionary definition of socialism is, rather than a particular interpretation or lived experience of something called it, it's "(a political system wherein) the means of production, distribution, and exchange should be owned or regulated by the community as a whole".
The governmental decision to offer something for free to everyone is a socialist one; the use of government regulation as an implicit threat to pay what is reasonable to offset cost, rather than what the market could dictate, also speaks to socialism. Yes, it was a private company that produced it, but both the exchange and the bulk of its distribution are very much being carried out by government, and even where not, are being very tightly regulated by government. All of that in response to community need and desire, not merely the strictures of the market (which otherwise would have seen a major profit opportunity).
Appreciate your comment, it’s good learning for me to hear a different opinion.
I might have extrapolated the benefits of a capitalist system too far.
But the main point still stands I think, the response to the above comment and the fact that it is very much those strong incentives that drive much of the innovation in our society today. It is not someone forcing someone to do something or pure good will that lead to these results.
That I can agree with. Ensuring people can reap some benefit to their efforts is important. The problems with capitalism that I personally have tend to be around places that people are able to reap benefit from other people's efforts unduly (i.e., billionaires), or from other people's needs unduly (i.e., shareholders making bank from healthcare).
> As an example, Capitalism gave us a vaccine to a novel corona-virus within a year. These companies didn’t do it out of good will, they did it because they knew they could make money. Many socialist countries are struggling because it turns out it’s hard to force people to produce innovation without the right incentive structures in place.
Most vaccine development occurs in academic research institutions, even if the large-scale manufacturing is then taken up by the capitalists. For example, the Covid vaccine manufactured by AstraZeneca was created by and innovated upon by a vaccine research group in Oxford University, who are obviously not in it for making vast wads of cash. And in general, it's the state who funds vaccine development, and the state who buys the vaccines.
On the broader subject of medicine, capitalism also gives us harsh enforcement of supposed 'intellectual property rights', stymieing the availability of medicines in developing countries. Fortunately, it looks like this will be waived for Covid vaccines, thanks to some government intervention. But it shouldn't have to be this way. Jonas Salk had the right idea when he released his polio vaccine freely to the world, to benefit all, without profit motive.
> As an example, Capitalism gave us a vaccine to a novel corona-virus within a year.
You realize Cuba also produced a vaccine within one year. So did China; though some people consider Dengism a kind of capitalism, so who knows, good for you, maybe.
All governments are slowly and inexorably moving towards a synthetic mix of capitalism and socialism. The problem in the US is _unfettered_ capitalism, which has metastasized into an attempt to capture the ENTIRE vertical stack of any business enterprise. That's the whole play of VC money today: find a segment of the economy to monopolize, run everyone else out, and then extract ALL the profits, from top to bottom. In my opinion, we need GDP-adjusted limits on how large a company can be, in several different dimensions.
And the fact that he states that he uses the word in the right way doesn't make it right. In fact, I find the use of the word paradigm quite shallow and disconnected from the Kuhnian definition.
>And the fact that he states that he uses the word in the right way doesn't make it right. In fact, I find the use of the word paradigm quite shallow and disconnected from the Kuhnian definition.
Perhaps because Kuhn doesn't dictate the definition of paradigm (and in general, the originator, etymology, etc. doesn't dictate use), the language community (the speakers of a language) does.
The whole SV/VC thing is just a big shallow marketing scheme to get young people buy into bad deals because they are brainwashed into thinking it's the cool thing to do. Have been there, done that, regret it. It took me years of being outside of SV to realize how brainwashed I was by comparing myself to the media and people around me.
> It took me years of being outside of SV to realize how brainwashed I was by comparing myself to the media and people around me.
The media doesn't represent normal people, and there are different regions of the country with very different norms, cultures, and beliefs. I know we are social animals heavily influenced by those around us, but it's also important to reflect and try to determine if those influences are right or productive.
Capital is the most conservative thing that exists. Any little sign of trouble and it flees. VC is not different, they just play the odds and invest in a way that the money is guaranteed back, but the optics is that they are "taking risks". They then lobby to have the goverment take that risk away, lobby against employees having power to never have any risk and so on. Capitalists are the most risk averse people.
"Crazy new ideas"? A VC, Graham, is praising crazy new ideas?
After I sent some hundreds of emails to Silicon Valley, NY, and Boston venture capital firms with nearly no positive feedback and nearly no feedback at all, I concluded several points:
(1) VCs won't invest in, consider, look at, or pay attention to crazy new ideas. Might guess that one reason can be that when the ideas are deep technically the VCs don't have the expertise to evaluate them, but the VC rarely seek evaluations from technical experts either. Net, VCs don't want to invest in crazy new ideas and hardly value ideas at all.
(2) My best guess is that most of all VCs like to invest in traction already significant and growing rapidly. The traction most desired is after tax earnings, but also good enough can be pre-tax earnings, revenue, or just Web site traffic.
(3) VCs also like the traction to be in a big market.
(4) VCs also like a team of several founders: Maybe the VCs are afraid that a sole, solo founder would get into human relationship problems as their company grew, and the team of several founders helps alleviate that fear. Also with several founders, if the VCs don't like the CEO, then the VCs can fire the CEO and promote one of the other founders to CEO.
E.g., I have had at least two crazy new ideas for new businesses:
The first idea was new and much better than anything else for real-time monitoring of health and wellness of servers and networks. I had running code and some quite good results on a variety of real data. The idea would not have had the potential of building a company worth $10 billion but might have built one worth $500 million and, maybe, with more advanced versions of the product, continued to grow. I gave a talk at the main NASDAQ site in Trumbull, CT, but apparently no VC had any interest at all. So, I went ahead and published the work; so, at least it was good enough to pass peer review!
Second I have an idea, and 100,000 lines of .NET code apparently ready for at least initial production, for a huge market and that, if people like it at all, and there is various evidence they will, should be worth $10 billion, maybe 200 times that if I further develop the work and people around the world like it a lot, and there is some evidence they might.
The idea makes powerful uses of some poorly known and understood advanced pure math (maybe understood by fewer than 10 computer science professors in the whole world, and only a tiny fraction of pure math professors will look for or see the connection with computing or business) and some original applied math I derived likely beyond over 90% of computer scientists.
But apparently no VC in the country is interested at all. Same for YC, the NSF IIP, etc. But, and really I designed this project this way, I don't really need funding now and won't if the traction grows. I've had some delays from unpredictable outside interruptions but am about to return to the work.
I can believe: The $500 million is too small for VCs to care, and the 200 times $10 billion is too big for them to believe.
> I can believe: The $500 million is too small for VCs to care, and the 200 times $10 billion is too big for them to believe.
This is actually true. The model of VCs requires that they take extremely high risk with extremely high upside. But their job is to derisk as much as possible. They derisk market risk, technical risk, and operational risk.
That is kind of reaching for the stars. Or shall I say last straw argument.
I think a much easier winner is if you have a domain expert in one area and bring his or her knowledge to another field.
A Designer who creates a web app to design logos. A fitness guru who develops a Drink. A car maker who builds space ships.
The reference to the Copernican Model of the solar system is an interesting one since it produced less accurate predictions than the Ptolemaic Model. Improving the predictions of the Copernican model initially involved the introduction of epicycles, diminishing the value of the crazy new idea.
It took the work of Kepler (elliptical orbits) and Newton (a physical basis for elliptical orbits) to elevate the heliocentric model to the status that it enjoys today.
There are two reasons why I bring this up: one is the validity of many of Paul Graham's assertions and conclusions. The other is to point out that things aren't so simple. Copernicus did not reap the rewards of his ideas since it took the work of others to prove those ideas. In fields outside of science, there is little reason to expect people to arrive upon similar conclusions. (Even within the sciences, there is no reason to believe we would converge on similar conclusions in the same time frame via different paths.)
Graham has a tendency to use hindsight to show that, yes, indeed he is contrarian and genius.
The Ethereum people do it, too. They use the example of old massively successful tech to prove that Eth too can catch on, but they don't cite all the failures.
Paul Graham is high on his own ego. Unfortunately, too many people take what he says at face value, like he's some sort of magical preacher of the Valley.
Please don't cross into personal attack on HN, for any meaning of $person. And please stop creating accounts to break HN's guidelines with. You're damaging the community you're part of by doing these things, and that is self-defeating behavior—especially if you consider how fragile this place is. I'm sure you're not the kind of person who would drop lit matches in a dry forest or even litter in a city park. Please stop doing the equivalents here.
The odds are high that the internet converges to suckage. We're spending a lot of energy trying to stave that off. Since you're such an active community participant, wouldn't it be in your interests to help that effort, rather than hurt it?
I think there's a very strong analogy between the Copernican model vs Ptolemaic and computer languages (for example) - you can have something like C which has obvious defects (grant this for the argument, substitute something else if you like), but a replacement like Rust isn't taking on just C, but the entire ecosystem around C and all the tooling and knowledge therein.
Similarly the Copernican model was "more correct" depending on how you look at it (currently we model everything relative to everything else, the earth and sun "orbit" around a midpoint that is inside the surface of the sun but not the exact center, for example) but it provided WORSE predictions than the Ptolemaic model of the time.
And this had real practical implications in the technology of the day - navigation charts, etc, analogous to trying to use new tooling and finding it doesn't support aspects the old tooling did.
He lists two reasons why people want to dismiss "crazy new ideas": envy, and the desire to seem sophisticated. Using Copernicus as an example simply does not fit these claims.
Reasons to dismiss Copernicus's ideas:
1. They were worse predictions, so for all 'practical' purposes of the day (essentially, only knowing where the wandering stars would be), adopting his ideas would be poor.
2. Most crazy new ideas like this proposed by people on the fringe of a field are flat out wrong. The Copernican Model is a good example of a common mistake, really: Someone saw a complex but accurate system and tried replacing it with a simple system that is easier to understand but rejects the actual data. Turns out the world is complex and doesn't care about being intuitive. See all of quantum mechanics for an example.
3. Classical relativity was only established later by Newton. The moon, sun, and other planets were poorly understood at the time. It wasn't until Galileo's observations that evidence was gained for a rocky moon. The idea that the whole Earth could be moving and spinning without violating everyday observation, and therefor the other bodies in the sky follow the same rules as those on Earth, is simply a large and unnecessary leap in intuition. Sure, it's a correct leap, but the path of reasoning there is backwards. It is only with Newton's laws of motion that heliocentrism begins to make any sense.
4. Further, much better evidence (as discussed above) came to light far before Copernicus's ideas had any effect on non-academic matters. Other than finding things interesting, and generally liking to know how the universe works, heliocentrism has no practical affect on life even today. Sure if you work at NASA it's super important, but most people don't. Trying to force an idea before it's time, when it won't effect things anyways has little practical value.
The crux of this is that this article is not a response to people dismissing a new idea. It's people dismissing a new business. He seems to be arguing that new ideas by domain experts shouldn't be criticized because they might be right, while ignoring that people can have strong financial motivation to promote incorrect ideas. Fighting such ideas is good because:
- If the ideas prove correct, internet criticism of it really won't matter. Only a few key investors need to be convinced, and yeah they'll make a lot of money.
- If the ideas are wrong, healthy skepticism is the strongest force against snake oil salesmen.
90% of starts fail, so the criticism will usually be on the correct side, even if the exact criticisms don't point to the true cause of failure.
> heliocentrism has no practical affect on life even today
I disagree. These discoveries (showing that we aren't "special" or at the "center") had huge religious/philosophical implications and through some intermediate steps ultimately led to the kinds of secular states we (most of us) spend our every day in practice.
should be prefaced with "within the context of american capitalism", because i don't think such views on personal ego and taking credit apply universally. There were a lot more crazy ideas when experimental science wasn't there to debunk them. And many science fields even today are lacking ideas, not methods, but not necessarily for fear of being ridiculed.
I feel like the same thing can be said about Boring Old Ideas. If the person who is speaking to you knows what they're talking about, there's good chance that they've identified a gap/opportunity and you should listen to them.
This seems more like an exercise in listening and trust than whether the idea is crazy or novel.
I do not think that the problem is with the content, moreso with the unnecessarily aggressive tone. Shit, cult ... - HN generally rewards civility and does not like sliding towards Twitter mode of operation.
Imagine if people could completely change their idenitites every time they said something. It would make conversation and discourse as we know it impossible.
It would make the idea you created an account to express invalid. You can't criticize Paul Graham if he can say "That wasn't me, it was Gaul Praham".
If your ideas aren't good enough for you to stay by them with some form of consistent identity, do not express those ideas. Think about them, develop them and then communicate them when they no longer embarrass you.
I recall "Dr. Gene Ray" defending the credibility of Time Cube on similar grounds. There wasn't enough of an idea presented to even have an argument about.
Fully agreed. A lazy article all-round. Where are the examples? Where are the pitfalls?
Is the insight merely that some wacky ideas turn out to be right sometimes?
Yeah, sure so how can we tell in advance? How is this actionable? And his answer is: listen to reasonable domain experts. That phrase is doing some real heavy lifting there...
Pretty much all of his articles are similarly lazy (edit: and usually articles from other top SV folks). I'm always astonished that people upvote his stuff. I've come to think of it as an HN tax: we get a well-moderated forum with otherwise good content and in return we have to scroll past one extra article every month or so.
All this stuff about Copernicus, Darwin, Galileo... You're not revolutionizing our understanding of humanity's place in the universe... You're buying companies that make filesharing apps and (often untaxed) hostel booking websites.
I agree that it fits here just fine (more than a lot of the crap that hits the front page) but I also agree with the OP that the “cultism” is annoying, not that I expect anything else given his association with the company that created HN.
pg is one of the people introduced me to lisp and i think lisp is the greatest idea in computer science. so his ideas have some weight for me, but lately the disconnect is quite unbearable. he is again picking the wrong side in history.
will these supervillians ever stop? nothing less than world-domination is enough.
microsoft tried to own everything about computing, they largly succeeded. they even infiltrated schools and still every pc is sold with windows-preinstalled.
then came google/facebook/apple... and they all want to own the entire internet.
unity wants the entire gaming/gamedev/computer-graphics. i am sure there are schools out there have unity classes, just unbelievable. "unity-developer" is the norm in job titles now, and i have never once heard "photoshop-artist" in the title of a job description.
This is a great essay. I follow similar logic when assessing bitcoin.
1. Listen to as many viewpoints as possible
2. Remove the people who are not domain experts
3. Remove the people who are blinded by their own financial interests or ego.
4. Remove the people who are influenced by social pressures ( good and bad )
5. Assign probabilities to the remaining viewpoints evenly.
6. Repeat steps 1-5 and chart the progression of the probability space.
Currently, It seems 40%+ likely that Bitcoin will end up as an accidental Ponzi scheme with a crash based on late adopters rushing out. But there is also still a 10%+ chance that it replaces gold as a currency hedge for the next 20+ years and 5+ percent chance that it becomes a first class currency.
I'm new at this, having just purchased some miners, but have followed the space using a similar logic. For the 10% view I'd consider the insane amount of institutional money chasing the entire value chain, which will not be shy about protecting itself.
The policy environment is still evolving; the FinCEN travel rule will put the hurt on crypto ATMs and the exchanges (to some degree). But it's good collateral, it doesn't (generally) have storage costs like gold, the growth of supply is well understood. Potential positive catalysts remain, including a US btc ETF, the impact of cme's micro contract, and geopolitical developments further constraining the supply of ASICs.
Its clear from the above that I view it more as a store of value than something you should spend as a currency. Ethereum-world (incl stablecoins) seems to have a lock on the latter. It will be interesting to see what happens when ETH shifts to POS.
Who exactly is a "domain expert" in bitcoin? The whole thing was created by an anonymous developer 10 years ago, and all of the so called "domain experts" are johnny come latelys who smelled a quick buck.
The whole thing was not created by an anonymous developer 10 years ago. The building blocks used in bitcoin have been around for many years before this in things like hashcash. Furthermore distributed consensus is a topic researched for decades. The term cryptocurrency uses crypto as a base which has been researched for centuries. Bitcoin is a novel constuction around crypto primatives relying on economic incentive to solve distributed consensus. There are clearly domain experts here.
Andreas Antonopoulos, Vitalik Buterin, any of the core devs at https://github.com/bitcoin/bitcoin, anyone who has studied blockchain and cryptographic currency systems for the past 10 years, etc. Keep the wool over your eyes.
> Someone proposes an idea that sounds crazy, most people dismiss it, then it gradually takes over the world.
This is survivorship bias. He is not wrong in that yes, we should consider ideas from domain experts carefully and not be overly dismissive. But it is absolutely survivorship bias to believe that every crazy idea from a domain expert worked out.
Graham should know better, he talked about arc as the messiah for five years before getting beat to market by Clojure. Now they barely do proper language releases.
He doesn't say that crazy ideas by domain experts always work out. He makes two claims:
1) That if the crazy idea is from a domain expert, it deserves to be taken seriously. It doesn't deserve to be taken as gospel, but it shouldn't be dismissed outright.
2) He's betting (but not stating as fact) that if you did blindly support the crazy ideas of domain experts, you'd be right more often than you're wrong.
This is not the same thing as saying domain experts are always right. Simply that all things being equal, you're better off betting on their expertise over your own ignorance. But even better than that is to look into the matter until you have sufficient expertise to properly evaluate their idea.
> Such ideas are not guaranteed to work. But they don't have to be. They just have to be sufficiently good bets — to have sufficiently high expected value. And I think on average they do.
It's likely that you're both right. If something is biased 45% off of a coin toss, but gives 25x returns, it's still positive EV.
This is a part of VC and speculative investment that a lot of people gloss over. Success is the difference between being wrong 95% of the time and being wrong 97% of the time. (And, with Softbank around, sometimes it's the difference between being wrong 97% of the time and 99.5% of the time.)
Having crazy ideas seems to be a prerequisite for taking over the world with an idea. The alternative is having familiar ideas which are already pervasive, or having no ideas.
It would be survivorship bias if the author suggested a rule here - but this quote is only a statement that 'this can happen'.
Where the author suggest a rule: "Such ideas are not guaranteed to work. But they don't have to be. They just have to be sufficiently good bets — to have sufficiently high expected value." - it is hedged against survivorship bias.
“But the fact that some geniuses were laughed at does not imply that all who are laughed at are geniuses. They laughed at Columbus, they laughed at Fulton, they laughed at the Wright brothers. But they also laughed at Bozo the Clown.” -Sagan
807 comments
[ 2.9 ms ] story [ 392 ms ] thread>3] This is one reason people with a touch of Asperger's may have an advantage in discovering new ideas. They're always flying on instruments
On the contrary, it is everybody else who is flying on instruments. The sperg sees straight.
The instrument tells them they are level, but they get paranoid and either stall or crash because they don't trust that the machine is more reliable than their emotions, even when it's drilled into them.
[1] https://medium.com/entrepreneurship-at-work/non-consensus-ba...
[0] https://www.mightyapp.com/
[1] https://twitter.com/Jonathan_Blow/status/1387101172230672389
[2] https://twitter.com/cmuratori/status/1387645578067124224
I'm sure there's a niche, though. Like low-paid workers needing to do a lot on their crappy personal machines.
Maybe there’s a market for people stuck on old computers but whose companies still spring for super fast internet and $50/month SaaS bills per user instead of just spending that same money (or less) financing the laptops they actually need, but it would be a small market.
Maybe even prevent browsers from downloading files onto the local computer. A full recording of each user's sessions. Integrated password manager that uses the Mighty login to tie them all together.
So I view it as valuable to corporations and thus mighty falls into the B2B category of company which makes software end users hate, but corporations love.
Probably a fair bit of money in that.
Not denying your other points but Amazon Workspaces[1] is a product that perfectly fits their needs.
From a top-level exec's stand point I would imagine they would be more willing to buy something like Amazon Workspaces which gives them 100% control and peace of mind Vs piecemeal approach such as browser, conference call client etc.,
[1] https://aws.amazon.com/workspaces
"Essay is french for attempt." There are good and there are bad attempts, of course.
(As we can see here, where a quarter of the comments are debating the business model of Mighty.)
Reading this I thought of CNIs like the Internet of the 1980s, women’s suffrage in the 19th century, Project Mercury, The Eiffel Tower.
The CNIs gaining ground today in tech seem to be crypto, brain-computer-interfaces, quantum computing, and CRISPR gene editing.
I think Mighty has unfairly taken the brunt of a lot of growing discontent with bloat on the web. I'm not ready to fork over $30+ for faster browsing, but I'm glad someone is working in this space, because I genuinely would like to rent a fast virtual computer for the occasional video editing task and think Mighty is a step in that direction.
Can't you already do that pretty easily on AWS or Azure?
What I’d like to see is a service that let me add my credit card, click a button, and launch a video editor on a high-RAM machine that I pay for the hour (including the software license fee).
My issue is that it’s not a product I could feel good recommending to anyone, at least at the high pricing that was proposed in the last discussion. In an era of $999 M1 Macs (and even cheaper AMD laptops) and readily available financing options, it doesn’t make sense for anyone to throw their money away at a SaaS service that simply cannot perform as well as local Chrome on a modern machine.
I could see the narrow use case for limited situations where someone has
1. Weird IT department restrictions that require them to use old, slow computers but also
2. Budget rules that allow them to spend monthly money on a SaaS but not on financing the hardware they need to get their job done and
3. Guaranteed high speed internet all of the time and
4. An IT/corporate security department that is okay with them sending all of their keystrokes, login info, and browser data to a 3rd party service
Surely this situation exists, but it still feels like Mighty is targeting a broader audience by providing untrue claims about remote thin client technology somehow being faster than a halfway decent local machine. I’d feel equally uncomfortable if Stadia was charging $50/month while claiming to be lower latency than local gaming.
The counter arguments about disrespecting hard working startup founders or doubting visionaries feel like a strawman response to legitimate questions about the value of their service. The technology and execution look to be good, AFAICT. It’s the product, pricing, messaging, and value that I can’t recommend.
How on earth would that work for their product? They need to access the data so they can render it.
Not to mention, it's not like Chrome can run Javascript without looking at the data, and if they're not running Chrome, they'll start getting into compatibility issues.
Sure, it's the ultimate sandbox for the code itself, it probably protects me perfectly from Meltdown attacks against other programs on my main system, but I don't think that's enough to call it 'endpoint security'.
So they need to figure it out otherwise it's only valid for personal use, which is not exactly the money is, imho.
Maybe some self hosted solution for corporate departments? Or waiting for an acquision from Citrix or VMWare...
Or just an IT/security department that is blissfully unaware of the shenanigans that some of its users are up to. Especially if people are accessing this from outside of the corporate network to start with.
My company uses webmarshal which fronts requests and blocks pages deemed not good for productivity/data security. In that type of situation you already have an intermediary so mighty could make sense - make everyone use mighty and give it a custom blocklist. Obviously not a situation that employees would be a fan of, but it’s something that corporate is already doing, and mighty could improve the experience.
To me Mighty sounds like a similar category of product. They are betting that PC/Laptop/Mobile hardware will stagnate from this point on. Exactly when Apple has launched M1, which blows the previous version out of the water, at a non crazy price. From this point it’s a matter of time other hardware also catches up in terms of performance and price.
Besides, yet another company to handover my entire browsing history and data for purported improvement in latency? I don’t know.
[1] https://en.wikipedia.org/wiki/Webaroo
Me playing the devils advocate...
Perhaps not //forever//, but perhaps for the next few years (enough for them to make some good money).
Why? Chip shortage, GPU shortage.
Anecdotally, I just bought a computer for $7500, that, two years ago would have cost about $2500 (granted, it is the new tech, but even so, new tech of this tier two years ago would have been approx. $2500). On top of that, have to wait two months for it to be assembled. (ouch)
I agree with you that speed will probably not stop increasing, but if prices continue to go up, or even just level off, few people will be able to afford the new tech.
That being said, would I invest in Mighty? No since I agree with you in general....and agree with you about the zero privacy issue...very regressive position for someone like PG imo.
What if, when you purchase your machine, it includes a host of features including a could suite of things like; VPN, Cloud Browser, An actual amount of decent cloud storage, portable applications, hosted in your cloud bucket, but accessible from any device (If I have Illustrator, I can just run it from any machine - not just my own. I can allow guest access to my paid licenses - like Say i want my brother to be able to draw stuff on my Illustrator license while I am not using it.
Etc.
Imagine if instead of that $7,500 "computer" you bought a $7,500 'stack' and the access terminal you happen to be using most is the physical laptop that youre used to.
I tried to build something similar to this more than a decade ago (2007 or so - but wrote about it in 2004).
My biggest issue is better information/knowledge management.
I have TBs of files and data strewn all over. We should be focusing on "my information" and let me manage and secure that - and access it from any device easily, and securely.
However I think we gonna do have RTX in our mobile devices - AR glasses, whatever modern mobile tech - because of decentralization (security/ethics purpose)
No, they are betting that no matter the processing power of computing devices of the day, developers will always push things to the limit and ship products that can end up being slow on devices used by a large share of the population.
This has always happened, because software product makers always want their products to have as many features/capabilities as possible, and release them to market as quickly/cheaply as possible, which means there are lesser incentives to limit functionality or to invest in performance optimisation.
There's no reason to believe this trend will cease, so Mighty is offering a service that allows people to get much higher performance of their webapps without always needing to have the highest-power computing device in their possession.
Agree, it's an arms race at this point between end consumers and app developers. However, it also means once Mighty becomes popular they have to contend with developers overwhelming their computing resources as well.
Mighty can succeed yes, but it's success would be bad for the ecosystem.
Centralized processing of all computing, would just make current issues like censorship, security issues worse.
No, they can't. Only the right-tail power users will ever pay for this. Unless almost all of a product's users are in that group, the developers still need to offer satisfactory performance to the rest of their users, or be vulnerable to a new competitor.
> Centralized processing of all computing, would just make current issues like censorship, security issues worse.
People keep saying this. I doubt even Mighty believes they or their paradigm will take over "all computing" - just the niche that really needs it. Even a small market share can make them a big/successful company, but without making the whole internet significantly more centralised than it already is. I don't get the panic.
All that said, it's worth noting that the first set of customers seem super impressed. So maybe Mighty are on to something.
That should easily support twice the peak CPU power for me when I need it.
Another thing is that the Mighty CPUs will (presumably) be upgraded continuously, while my laptop CPU gets no faster after purchase. If that makes me only buy a new laptop every 4 years instead of 2, I've saved a lot of money and hassle.
I'm not saying this means Mighty will conquer the world. But there are reasonable arguments behind the model. Especially if you assume bandwidth will keep improving.
It's not. It's literally spinning up a VM in the cloud to run Chrome and stream a video to you
> At least 80% of the time my computer isn't even used.
Yes, but Mighty isn't running (and will never run) on your computer.
Well, it most likely shares the CPU with other VMs, but that depends on the cloud, and the instance. And since the browser is always open, Mighty will always run that VM, with no sharing: https://news.ycombinator.com/item?id=27063554
So the idea is you can get your laptop utilization down to 1% by pushing your load to their cloud? Fascinating.
Yeah, but I think it's better to think about it like virtual/remote desktops, where the granularity is a browser tab.
At least, that's the only way it makes sense to me. I might understand that to do 4k high res 3d on my phone - my phone will need help. And once that "help" is available - my TV, and my tablet and my smart watch can make use of it.
But I'm not generally willing to trade latency for cpu time sharing. What is interesting is an always on, always working desktop session. It's why I like screen/tmux and ssh, rdp - and would consider running a Linux terminal server, so my laptop(s) and desktop(s) could be a disk less, stateless thin client.
Make the observation that the browser tab is the new process/application granuality - and it makes sense to host tabs in the cloud.
Personally I'd want to self host it - but the idea doesn't sound quite so inane.
Not only that but a great amount of people here live in a literal tech bubble and we believe that everyone has a reasonably fast laptop or can setup a large workstation.
By that same token the average persons browsing habits are also not the same. Optimising for Chrome power users who have eleventy billion tabs open and run hefty web apps isn’t the normal usecase. My families browsing is mostly taken care of on a 2012 MBP and a couple of ancient iPads. Mighty would be functionally useless for us.
Heck my entire day is spent in Chrome working in Drive and writing code in our own development environment in our web app. I am a browser power user in that sense on a three year old gaming laptop and I don’t understand the Mighty usecase. It’s a niche within a niche at the moment although I understand the appeal of the business model.
I’ll admit, my initial reaction to Mighty was “I can’t imagine it ever being faster than my behemoth PC.” But then I stopped for a second and got curious. What is Mighty, really? It is a thin-client. That’s it. And are thin-clients a bad thing? Well, if latency is an issue, sure. But “high-ping” is a somewhat solved issue, whether in multiplayer gaming, in terminal utilities like Mosh, or even in optimistic GraphQL mutation updates. Where’s the use-case where near-zero latency is vital? The only cases I can think of are games like Rocket League: fast-twitch games where even the latency between my controller and my PC is something I happily spend hours optimizing — where latency prevents the necessary feedback loop for learning (akin to trying to learn how to hit a baseball while drunk).
But beyond near-zero latency use-cases, why would a thick client ever be better than a thin client? At the edge of performance, this question is easily answered: I would never attempt to train a PyTorch model on my admittedly powerful GPU. That’s what the cloud is for. So when it comes to my browser, why am I content to eat up memory and cpu-time with hundreds of tabs open that almost always include one or two that are broken, soaking up my resources, and have to be hunted down and killed off so that IntelliJ can return to its normal lightning-fast speed?
Might goes even further and asks why I would want to run IntelliJ on my machine at all. Wouldn’t I rather run IntelliJ like I used to run Vim over Mosh, where I never have to worry about storage space, about download/upload bandwidth, or about my computer becoming sluggish?
And that’s the killer idea here: that thin-clients almost always beat thick-clients. One could even argue that the entire internet is premised upon this reality.
I’d happily pay Mighty to try it out for a bit. Even if it doesn’t work, I’ve dropped more money of less fascinating ideas. At the very least, I’m rooting for their success, because it would change a lot more than how you consume content over the internet.
Edit: Managed to find the right string of words to google: mosh = Mobile Shell
The biggest gain here is security, you won't ever run code natively. The caveat is you also have to trust the host.
That assumes any malware cannot bust out out of mightyapp's sandbox to the host's host.
The biggest loss is also security: all of your keystrokes, passwords, traffic, etc. are stored on someone else's hardware.
And as you note the need for more power goes beyond a browser, IDEs, gaming, rendering, Bitcoin mining. Why just do it for the browser? You can, and some people do, remote into a more powerful machine for all of their work. This was the norm when terminals were true terminals. We could go back to this, but having your own computer historically had much larger benefits for people.
Because thick client distribute the load among many computer. With a thin client all that load is way more centralised.
> I would never attempt to train a PyTorch
No one is asking you to do that. We're talking about Mighty, a thin client that:
- runs a beefy VM in the cloud
- runs a single app, Chrome, in that VM
- streams video to the client
If a million people run Chrome on their laptops and keep it open for the entire day (and your browser is usually open throughout the day), that's... just a million people with their laptops.
If a million people run Chrome through Mighty, Mighty needs a million VMs always open, and a million video streams, also always open.
See how a thick client is better than a thin client?
I was curious about Mighty and looking forward to their technology. That’s not the problem.
The judgment came largely when they announced that it cost up to $600/year. It costs so much that it’s actually cheaper to buy a whole new computer if you might need it for a year or more. Once you put a price tag on something and ask people to pay for it, judgment is fair game.
This whole dismissal of people saying that they couldn’t justify the product for the price as some sort of anti-curiosity thing feels disingenuous.
His arguments are too vague to specifically defend Mighty. You can insert any technology and his arguments are neither valid or invalid.
I would say typing is a pretty big one. It is extraordinarily unpleasant when typing lag is anywhere above maybe 50ms,and even worse when it is variable, like it would inevitably be if going over the Internet. It's even worse with mouse movements, where occasional spikes in lag can ruin your day.
Not to mention, the problem is fundamentally simpler in a terminal, with a very limited range of outputs. The problem of optimistically drawing the result of your input on the screen is much harder when that input could affect any portion of the screen in any way, like it can on a JS-powered web page.
I haven't found Chrome slow on my mid-range PC (i5-7500). Or my phone (Pixel 3).
I feel like I must be in some parallel universe to everyone else talking about how slow Chrome is.
Just now, I loaded the first 6 links on Hacker News. 1 didn't load at all due to a server error. The other 5 all loaded in under a second (measured by DOMContentLoaded). I have uBlock Origin enabled (only in Firefox on the phone). Maybe that helps.
I can have 100+ tabs open without slow down if I want to. The bigger problem is I'm less productive with 100 tabs open because... there are 100 tabs open. It's just too cluttered.
I'm willing to admit Mighty might be a good product if people have this problem with slow browsers. I just never found this was an issue. Maybe it's a problem on low-end machines, but how many people have a low-end computer but are willing to pay $50/month for Mighty?
I just set up a new computer and I was wondering why Chrome and Firefox was so slow. Then I remembered I hadn't set up AdBlock yet.
I wonder if Mighty has ad blocking? It's an interesting question of which will be better for their business -- blocking or no blocking. (I won't try it because of the obvious privacy problems, because I use a fast computer, and avoid slow web sites. But it probably has an audience.)
PG wants you to think that they're mutually exclusive. They're not, but he has a product to hustle.
The problem (for PG) is that curiosity is not uncritical. Curiosity poses more questions than those in the category "how well does this product work?" Anyone actually curious is going to wonder about the problem space, not just one proposed solution.
Pretty obvious questions include: "how did software get us to the point we're exploring this as a design?" and "could the problem it seeks to solve be addressed in a way that eliminates assumptions about the solution space?", "What are peripheral ramifications of design decisions, and how much do I care?" "Would other approaches solve the same issues, have the same ramifications?" or "Are they synergies to leverage by trying multiple things in concert?"
Here's a really good talk by him that dives into it, titled dramatically "Preventing the Collapse of Civilization": https://www.youtube.com/watch?v=ZSRHeXYDLko
Also note he's currently developing a programming language called JAI that's trying (and seems like it might ultimately succeed) in being a faster and smoother alternative to C++ for developing games.
> "50+ tabs without your computer coming to a crawl"
On Macs, people can just switch to Safari for free and solve that problem. Yes, Chrome is a memory hog. Stop using Chrome, don't send all your browsing data to a third party.
Perhaps Windows would be a better choice for Mighty demo shots, since there may not have a better option than Chrome for Windows.
At our startup we have a small org of customer service reps. They basically live in four tools: Notion, Slack, Chrome, and Front.
These four tools have one thing in common: they are all Electron apps. They SUCK. On any windows computer slower than a $1.5K lenovo, our reps can't have more than ~10 tabs open before their computer starts to stutter.
One answer would be to buy everyone an M1 mac. However, most of these reps are not familiar with apple (they use PCs at home) + they are still a bit too expensive (we're in Mexico where they cost 1.5 - 2x).
I would love to be able to buy our customer service reps a setup (monitor + mouse + computer) for ~$500 USD and have them use Mighty to run our customer service suite for $10-$15 bucks a month. That would scale to a customer service org of hundreds.
At $50 bucks a month I should just buy them a mac.
IMHO Suhail's pitch about "running figma" is wrong. I'm happy to buy a designer any computer they want. But outfitting a CSR team of 50+ people? Mighty + a thin client would be amazing.
Yeah, agreed. I kind of thought the savvy business plan would be to charge at a loss, make up the loss with VC money, gather a lot of market share (customers like you), then raise the price once the pain of moving off is too high
I don't buy the argument that performance will always get tapped out by developers. There is an upper limit to what even a horribly architected 2D web app needs to consume.
They didn't reveal their true business model or pricing yet.
The $50 subscription is a way to simply not flood their servers while beta-testing them, and also getting a bit of money from rich/dumb early adopters.
Fake it until you make it is still a thing in Silicon Valley.
Edit: Although, I'm not sure it supports multiple separate logins.
No, I'm upset that Mighty is necessary. That we, as an industry, have failed so hard that multi-gigahertz mult-core machines with gigabytes of ram can no longer consistently and quickly render documents without offloading it to a central server.
https://www.youtube.com/watch?v=pW-SOdj4Kkk
Definitely highlights something I've always felt, which is that we're really making life complicated with all those tools and processes that are less than ideal.
I get the feeling business and the need to "ship" stuff is the thing really bringing technology down, even though it looks like it's moving it forward.
I hear a implausible sounding ideas all the time. I think one of the main points of academia is to give people the chance to explore those ideas. But they don't turn out good "on average", not even close.
But if you never bet on any implausible-sounding ideas, you exclude the chance of being an early participant in any big paradigm shift.
A good exercise is to back-test your rule against big ideas through their history. Would you have invested time in solving Schrodinger’s weird equation in 1925? 1927? 1940? Would you have invested time in public key cryptography in 1975? 1976? 1980? 1990?
People who got in early, got to make the big discoveries in those fields.
Everyone assumes they would always have been on the right side of history. It's basically impossible to know what you would have been like had you been born in those times. Better is to back-test your rule against paradigm shifts that have occurred in your own lifetime. What are some ideas you dismissed that turned out to work? What are some products or projects you have a revulsion against that are still succeeding years later? What caused you to reject them? What did you miss?
But then they’d have to own up to the nauseating idea of callously dismissing some truly spectacular sea changes, such as failing to buy 1000 Bitcoins for a few dollars in 2009. I think most people are too full of envy to do that.
How many can honestly say they kept an open mind to the handful of such paradigm shifts we’ve experienced in the last 20 years?
So regretting past mistakes is a big waste of time. Learn from them, but don't let it get to you emotionally. (All easy advice to give, and hard to follow.)
Looking backwards definitely biases you to seeing things you didn't see at the time. Even memories get tinged. But memories are probably the most accurate you can do.
I missed bitcoin entirely despite knowing many people who bought as I still don't understand what the appeal is. It doesn't seem efficient to me as a payments technology or from a privacy perspective. I kind of wish I bought early even though I probably would have cashed out way too soon, as the money would have been nice.
I liked Stripe due to my revulsion to PayPal but wasn't exactly going to be on them winning. I was pleasantly surprised. Ditto Shopify because of my dislike of Amazon stores. I think I underestimate the chances of good ideas and good products winning because I know it's very hard.
I took a few bets early in my career on start-ups. I think I underestimated the difficulty of selling into retail on one company that I still think had a good idea. The others ended up being acquisitions, although the products weren't continued. I still believe social notetaking is an interesting hard problem in need of a good solution but I don't think we had found it.
I think the big lesson for me is to not get taken in by the cult of personality around strong founders and focus more on the merits of the ideas. That can be tricky as you're not getting the full picture when you interview making it even harder to see a crazy idea as good for the right reasons but it's the right approach.
Ordering pet food over the internet in 1995: crazy moonshot weirdo stuff.
Ordering pet food over the internet in 2021: just another day.
I initially dismissed the WWW as a disorganized mess when I first encountered it in 1994 - I liked Gopher better. By 1997 I had changed my tune and was all-in on learning HTML and CGI and even some of the new technologies like Java applets, RealAudio, and DHTML. (Note that Java and RealAudio were themselves eventually losers, but by then I was all-in on Javascript.) It felt like I had missed the boat in 1997; the dot com boom was in full swing and people were getting rich off much more sophisticated stuff all around me (TBF, I was in high school). But the WWW and Javascript/DHTML remained a lucrative career for 20 years afterwards.
A lot of people, once they take a position on something, dig in and don't change that position for a lifetime, then miss out on the opportunity of a lifetime because they dismissed it to begin with. But truly world-changing ideas usually have a lot of room to grow. If you're wrong for 2-3 years and then change your mind, you'll feel really stupid, like you totally missed the boat, but that's nothing compared to the 20-30 years left of growth that the idea might have.
In other words, a successful business requires not only producing an innovative product, it also requires knowing if people actually demand this innovative product. One half is the core idea, the other half is how this idea interfaces with the world.
Y Combinator and Tesla were seen as stupid and crazy for a long time. Meanwhile I’ve seen a lot of domain expert entrepreneurs do really crazy things that lead nowhere when it’s VC money instead of theirs.
iPhone certainly not the first touch screen display, Microsoft did have one, I believe. However Apple's timing was perfect, there was increasing adaptation of wifi connectivity and 3g, making it able to browse anywhere. Of course smoother interface and beautiful design also contribute big.
AmongUs was lucky they got advertisement from streamer, at perfect timing that was (is) pandemic and lockdowns.
Tesla can be considered crazy because they try to emerge with unknown product, with unknown market and unknown possibility of success. If we look at the past, Apple's iPhone was crazy too, because it was new product, with unknown market.
So it was a no-brainer for those already carrying an iPod and a phone, even if it didn't do anything beyond being an iPod that could receive calls and texts!
Timing would be hard, but I could think of some general ideas (behavior changes when theres a pandemic, assuming certain trends will go forward, etc)
> Most implausible-sounding ideas are in fact bad and could be safely dismissed. But not when they're proposed by reasonable domain experts.
Off the top of my head, I vaguely remember that the latter half of the careers of many "greats" like Einstein, Fermi, Erdos were riddled with pursuing directions which bore no fruit (to this day).
I don't quite think that crazy bets work on "average". On average they fail. It mostly so happens that crazy bets are on "average" taken by people with a network that can attract able and passionate people, who can see through an idea to its conclusion. This is where experts come in.
Experts often rely on validation by their "community", and by the nature of community dynamics need to upsell. They naturally attract the kind of people needed to execute crazy ideas. But then attracting such talent also has a corrective effect on the originally crazy idea, such that it has a higher chance of succeeding by minimizing the blind spots, the unknown unknowns.
I don't know about Erdos, but the other had ambitious ideas, not crazy ones. Evidence that they weren't crazy is that the problems they were working on are still seen as the ones to solve by most of the physics community.
In fact, on the case of Einstein, he worked on much crazier ideas on the beginning of his career. I don't think any respected physicist ever thought about solving Brownian motion, and the photoelectric effect was an interesting curiosity but mostly of interest of "those inventors out there", not of researchers.
Black swan events can probably describe a lot of things, but my understanding is that an individual or a team does not really set themselves up for black swan events. It just happens.
And don't forget Newton!
That's the VC business model in a nutshell
Google didn't come out of Yahoo. Napster wasn't founded by music industry execs. Friendster didn't evolve into Facebook. Match.com didn't invent Tinder. It's often the curious outsiders that disrupt the status quo, not the well connected insiders.
Ditto Xerox Parc and modern user interfaces, ethernet and and laser printers.
You won’t find a revolutionary import/export company that doesn’t have some knowledge about the logistics industry. But you are more likely to be paying attention to new technologies as they emerge that might allow you to come up with the revolutionary ideas. I’ve noticed as I’ve gotten older I’ve stopped spending my time fiddling with all my devices. As a result I’m less likely to realize when something new and exciting might be within reach, until it hits a higher point on the bell curve of maturity.
So Im skeptical of this outsider/insider characterization, IMHO it's almost always the individuals.
from my reading of the history of Yahoo they had plenty of good ideas that they destroyed because the problem was not domain experts but thinking they were in the wrong business (media!)
Less positively formulated, you could also say people outside an industry are naïve enough to start something people inside would deem impossible.
AirBnB, which is perhaps pg's favorite company, would never have gotten into Y Combinator without Michael Seibel and the endorsement of the JTV crew. AirBnB simply would not exist today.
So I hope people don't read this essay and think, "I'm not a domain expert, so I'll give up on this idea." Somewhat of a dangerous essay for that reason.
In the case of music, one could have domain expertise on audio files, compression and domain expertise about file sharing and domain expertise about younger people's listening habits. That's just having some insights into the domain, several as is the case with Napster. Music industry execs in those days probably had domain expertise on contract law, revenue sharing models, marketing and didn't have a clue about file sharing or audio compression.
Also want to mention the examples you mention seem to have not much "newness" to them to be "crazy new" in my opinion.
When I think of "crazy new" ideas, I think of things like stuff that came out of PARC in the 60's. If you go watch the Stanford 2 part series "How To Invent The Future" with Alan Kay on youtube [0] it goes into a lot of detail about them.
The challenge there with Xerox was they couldn't capitalize on every single idea that came out, just not logistically possible for a company to really do that.
[0] https://www.youtube.com/watch?v=id1WShzzMCQ
I don't know about that... look at physics for example. You have plenty of experts who are total mavericks and then a bunch of outsiders spewing crankery and messing up on the basics.
Larry Page and Sergey Brin were domain experts - they were doing a PhD dissertation on mapping and indexing the web.
The other businesses you mention aren't all that crazy; they just needed good implementation.
It's very difficult to differentiate between an innovative idea and claptrap. Genius is difficult to recognise because it takes a genius to develop the idea in the first place.
I think you are being overly pessimistic by several orders of magnitude.
My theory is that many domain experts haven’t needed to communicate with a lay audience in decades (if ever) and thus aren’t aware of their own baseline assumptions. Seems like a startup idea, maybe? Convert academic papers into comprehensible English. Two Minute Papers does this but it’s only for technology.
https://youtube.com/c/K%C3%A1rolyZsolnai
This probably applies to early Apple. Wozniak, while clearly a technical genius, needed Jobs’ communication and design skills to sell “personal computing” and make Apple a mass-market company.
Feynman loved explaining things, so he had to keep trying to explain them in a manner that the ordinary folk could understand, but he also loved physics, which kept him diving deeper, and playing with it.
Having charisma and knowing what people want, and how to sell was Steve Jobs portion of the game, once the initial technical hurdles were solved. Woz likes to minimize circuits and do clever things, that was his part of the game.
Here on HN, the balance seems to be keeping it focused enough on hacking technology, while still appealing to those who want to make money off of it in yacht loads.
It's all about balance between at least 2 domains.
What I find interesting where he calls most people conservative is how when I'm discussing VCs, YC, etc with non SV people is we talk about how conservative they are. YC's business model for years was predicated on finding talent (kids) that was undervalued, underpaying them (10k was the initial payouts?), getting them to pack up and move to super-expensive monocultural SV like everyone else, and then when they made bukku money, lose interest in improving the service (Reddit, Dropbox). There's nothing original there, it's the business model of carpetbaggers and robber barons. How boring.
I often feel like Jonah Hill in Moneyball, a pariah for pointing out how ancient VC thinking is. Or maybe I just imagine it. Well, it is my experience that true deep domain knowledge can only come from years of insights. People without years of experience will be lacking maturity and/or won't have time to even consider those insights. Usually people who mainly care about money, influence "becoming powerful" (Graham's words about his protege, not mine) will jump at whatever shortcut they can take instead of spending the card work necessary for learning these deep insights PG is interested in.
With all due respect, Coinbase didn't require deep insight. It was the equivalent of "You like money, too?" . PG, you should stop the contrarian persecuted intellectual look. You're not the little guy anymore, haven't been for 2 decades.
Indeed. Venture capitalism is still capitalism, with all its abuse and exploitation.
As an example, Capitalism gave us a vaccine to a novel corona-virus within a year. These companies didn’t do it out of good will, they did it because they knew they could make money. Many socialist countries are struggling because it turns out it’s hard to force people to produce innovation without the right incentive structures in place.
If you want to blame anyone blame God for making human nature this way. Capitalism is just the system that is most effective at tapping human psychology to push people to produce the best possible work.
Which can exist in a socialist economy.
As of July last year (again, let alone anything since), US governmental agencies had paid Pfizer $1.95 billion.
So, basically, the government spent billions for something they then distributed 'for free' to the taxpayers.
That's an interesting example to pick for 'capitalism'.
Just because it was the government that paid them, why isn’t that capitalism?
Socialism and Communism is when you tell people what they’ll charge for the greater good of the people. At least that’s what it is to me, someone from a former communist country.
Edit: not only did those companies get paid to invent it, but others got paid to make it, and yet others got paid to distribute it (pharmacies etc).
Really? I would have expected something a bit higher than $20-$25 a dose, given the limited supply (it was pure research at the time), and COVID was decimating the country (so extremely high demand). I mean, a single dose of insulin, requiring no research, and no protective IP, and costing cents to produce, will run you or an insurance company hundreds of dollars. Given the huge demand, and the non-existent supply at that time, there is -nothing- in the market that would have capped the price for the vaccine at $20-25, except either human good will (which is not a market factor; again, see insulin), or the very real threat of government action.
>> At least that’s what it is to me, someone from a former communist country
And voting Republican is what Democracy means to plenty of people living in the US. If we actually go to what the dictionary definition of socialism is, rather than a particular interpretation or lived experience of something called it, it's "(a political system wherein) the means of production, distribution, and exchange should be owned or regulated by the community as a whole".
The governmental decision to offer something for free to everyone is a socialist one; the use of government regulation as an implicit threat to pay what is reasonable to offset cost, rather than what the market could dictate, also speaks to socialism. Yes, it was a private company that produced it, but both the exchange and the bulk of its distribution are very much being carried out by government, and even where not, are being very tightly regulated by government. All of that in response to community need and desire, not merely the strictures of the market (which otherwise would have seen a major profit opportunity).
I might have extrapolated the benefits of a capitalist system too far.
But the main point still stands I think, the response to the above comment and the fact that it is very much those strong incentives that drive much of the innovation in our society today. It is not someone forcing someone to do something or pure good will that lead to these results.
Most vaccine development occurs in academic research institutions, even if the large-scale manufacturing is then taken up by the capitalists. For example, the Covid vaccine manufactured by AstraZeneca was created by and innovated upon by a vaccine research group in Oxford University, who are obviously not in it for making vast wads of cash. And in general, it's the state who funds vaccine development, and the state who buys the vaccines.
On the broader subject of medicine, capitalism also gives us harsh enforcement of supposed 'intellectual property rights', stymieing the availability of medicines in developing countries. Fortunately, it looks like this will be waived for Covid vaccines, thanks to some government intervention. But it shouldn't have to be this way. Jonas Salk had the right idea when he released his polio vaccine freely to the world, to benefit all, without profit motive.
You realize Cuba also produced a vaccine within one year. So did China; though some people consider Dengism a kind of capitalism, so who knows, good for you, maybe.
Perhaps because Kuhn doesn't dictate the definition of paradigm (and in general, the originator, etymology, etc. doesn't dictate use), the language community (the speakers of a language) does.
The media doesn't represent normal people, and there are different regions of the country with very different norms, cultures, and beliefs. I know we are social animals heavily influenced by those around us, but it's also important to reflect and try to determine if those influences are right or productive.
They are not in the game to fuel innovation, contrary to what they pretend.
They are in the game to protect old-money against innovation.
After I sent some hundreds of emails to Silicon Valley, NY, and Boston venture capital firms with nearly no positive feedback and nearly no feedback at all, I concluded several points:
(1) VCs won't invest in, consider, look at, or pay attention to crazy new ideas. Might guess that one reason can be that when the ideas are deep technically the VCs don't have the expertise to evaluate them, but the VC rarely seek evaluations from technical experts either. Net, VCs don't want to invest in crazy new ideas and hardly value ideas at all.
(2) My best guess is that most of all VCs like to invest in traction already significant and growing rapidly. The traction most desired is after tax earnings, but also good enough can be pre-tax earnings, revenue, or just Web site traffic.
(3) VCs also like the traction to be in a big market.
(4) VCs also like a team of several founders: Maybe the VCs are afraid that a sole, solo founder would get into human relationship problems as their company grew, and the team of several founders helps alleviate that fear. Also with several founders, if the VCs don't like the CEO, then the VCs can fire the CEO and promote one of the other founders to CEO.
E.g., I have had at least two crazy new ideas for new businesses:
The first idea was new and much better than anything else for real-time monitoring of health and wellness of servers and networks. I had running code and some quite good results on a variety of real data. The idea would not have had the potential of building a company worth $10 billion but might have built one worth $500 million and, maybe, with more advanced versions of the product, continued to grow. I gave a talk at the main NASDAQ site in Trumbull, CT, but apparently no VC had any interest at all. So, I went ahead and published the work; so, at least it was good enough to pass peer review!
Second I have an idea, and 100,000 lines of .NET code apparently ready for at least initial production, for a huge market and that, if people like it at all, and there is various evidence they will, should be worth $10 billion, maybe 200 times that if I further develop the work and people around the world like it a lot, and there is some evidence they might.
The idea makes powerful uses of some poorly known and understood advanced pure math (maybe understood by fewer than 10 computer science professors in the whole world, and only a tiny fraction of pure math professors will look for or see the connection with computing or business) and some original applied math I derived likely beyond over 90% of computer scientists.
But apparently no VC in the country is interested at all. Same for YC, the NSF IIP, etc. But, and really I designed this project this way, I don't really need funding now and won't if the traction grows. I've had some delays from unpredictable outside interruptions but am about to return to the work.
I can believe: The $500 million is too small for VCs to care, and the 200 times $10 billion is too big for them to believe.
This is actually true. The model of VCs requires that they take extremely high risk with extremely high upside. But their job is to derisk as much as possible. They derisk market risk, technical risk, and operational risk.
Get the traction, and the money will follow.
It took the work of Kepler (elliptical orbits) and Newton (a physical basis for elliptical orbits) to elevate the heliocentric model to the status that it enjoys today.
There are two reasons why I bring this up: one is the validity of many of Paul Graham's assertions and conclusions. The other is to point out that things aren't so simple. Copernicus did not reap the rewards of his ideas since it took the work of others to prove those ideas. In fields outside of science, there is little reason to expect people to arrive upon similar conclusions. (Even within the sciences, there is no reason to believe we would converge on similar conclusions in the same time frame via different paths.)
The Ethereum people do it, too. They use the example of old massively successful tech to prove that Eth too can catch on, but they don't cite all the failures.
The odds are high that the internet converges to suckage. We're spending a lot of energy trying to stave that off. Since you're such an active community participant, wouldn't it be in your interests to help that effort, rather than hurt it?
https://news.ycombinator.com/newsguidelines.html
Similarly the Copernican model was "more correct" depending on how you look at it (currently we model everything relative to everything else, the earth and sun "orbit" around a midpoint that is inside the surface of the sun but not the exact center, for example) but it provided WORSE predictions than the Ptolemaic model of the time.
And this had real practical implications in the technology of the day - navigation charts, etc, analogous to trying to use new tooling and finding it doesn't support aspects the old tooling did.
He lists two reasons why people want to dismiss "crazy new ideas": envy, and the desire to seem sophisticated. Using Copernicus as an example simply does not fit these claims.
Reasons to dismiss Copernicus's ideas:
1. They were worse predictions, so for all 'practical' purposes of the day (essentially, only knowing where the wandering stars would be), adopting his ideas would be poor.
2. Most crazy new ideas like this proposed by people on the fringe of a field are flat out wrong. The Copernican Model is a good example of a common mistake, really: Someone saw a complex but accurate system and tried replacing it with a simple system that is easier to understand but rejects the actual data. Turns out the world is complex and doesn't care about being intuitive. See all of quantum mechanics for an example.
3. Classical relativity was only established later by Newton. The moon, sun, and other planets were poorly understood at the time. It wasn't until Galileo's observations that evidence was gained for a rocky moon. The idea that the whole Earth could be moving and spinning without violating everyday observation, and therefor the other bodies in the sky follow the same rules as those on Earth, is simply a large and unnecessary leap in intuition. Sure, it's a correct leap, but the path of reasoning there is backwards. It is only with Newton's laws of motion that heliocentrism begins to make any sense.
4. Further, much better evidence (as discussed above) came to light far before Copernicus's ideas had any effect on non-academic matters. Other than finding things interesting, and generally liking to know how the universe works, heliocentrism has no practical affect on life even today. Sure if you work at NASA it's super important, but most people don't. Trying to force an idea before it's time, when it won't effect things anyways has little practical value.
The crux of this is that this article is not a response to people dismissing a new idea. It's people dismissing a new business. He seems to be arguing that new ideas by domain experts shouldn't be criticized because they might be right, while ignoring that people can have strong financial motivation to promote incorrect ideas. Fighting such ideas is good because:
- If the ideas prove correct, internet criticism of it really won't matter. Only a few key investors need to be convinced, and yeah they'll make a lot of money.
- If the ideas are wrong, healthy skepticism is the strongest force against snake oil salesmen.
90% of starts fail, so the criticism will usually be on the correct side, even if the exact criticisms don't point to the true cause of failure.
I disagree. These discoveries (showing that we aren't "special" or at the "center") had huge religious/philosophical implications and through some intermediate steps ultimately led to the kinds of secular states we (most of us) spend our every day in practice.
Humans are dogmatic about politics, the backfire effect is a thing, etc. Why should things be magically different for scientists?
There’s a reason “science advances one funeral at a time “
This seems more like an exercise in listening and trust than whether the idea is crazy or novel.
Just because Paul Graham wrote it, doesn't mean it's worth reading. Usually the opposite, in fact.
You people need to lay off the PG cultism. Stop worshipping every word he writes just because you idolize his riches, or want to be him, or whatever.
It would make the idea you created an account to express invalid. You can't criticize Paul Graham if he can say "That wasn't me, it was Gaul Praham".
If your ideas aren't good enough for you to stay by them with some form of consistent identity, do not express those ideas. Think about them, develop them and then communicate them when they no longer embarrass you.
you also got discourse backwards. you should be judging the arguments presented (which in this case not much), not the persona.
Is the insight merely that some wacky ideas turn out to be right sometimes?
Yeah, sure so how can we tell in advance? How is this actionable? And his answer is: listen to reasonable domain experts. That phrase is doing some real heavy lifting there...
That's a great way to put it!
All this stuff about Copernicus, Darwin, Galileo... You're not revolutionizing our understanding of humanity's place in the universe... You're buying companies that make filesharing apps and (often untaxed) hostel booking websites.
will these supervillians ever stop? nothing less than world-domination is enough.
microsoft tried to own everything about computing, they largly succeeded. they even infiltrated schools and still every pc is sold with windows-preinstalled.
then came google/facebook/apple... and they all want to own the entire internet.
unity wants the entire gaming/gamedev/computer-graphics. i am sure there are schools out there have unity classes, just unbelievable. "unity-developer" is the norm in job titles now, and i have never once heard "photoshop-artist" in the title of a job description.
this particular case is no exception.
ok, most logical conclusion is i too am envious.
1. Listen to as many viewpoints as possible 2. Remove the people who are not domain experts 3. Remove the people who are blinded by their own financial interests or ego. 4. Remove the people who are influenced by social pressures ( good and bad ) 5. Assign probabilities to the remaining viewpoints evenly. 6. Repeat steps 1-5 and chart the progression of the probability space.
Currently, It seems 40%+ likely that Bitcoin will end up as an accidental Ponzi scheme with a crash based on late adopters rushing out. But there is also still a 10%+ chance that it replaces gold as a currency hedge for the next 20+ years and 5+ percent chance that it becomes a first class currency.
The policy environment is still evolving; the FinCEN travel rule will put the hurt on crypto ATMs and the exchanges (to some degree). But it's good collateral, it doesn't (generally) have storage costs like gold, the growth of supply is well understood. Potential positive catalysts remain, including a US btc ETF, the impact of cme's micro contract, and geopolitical developments further constraining the supply of ASICs.
Its clear from the above that I view it more as a store of value than something you should spend as a currency. Ethereum-world (incl stablecoins) seems to have a lock on the latter. It will be interesting to see what happens when ETH shifts to POS.
Most BTC sits idle in exchanges like coinbase.
This is survivorship bias. He is not wrong in that yes, we should consider ideas from domain experts carefully and not be overly dismissive. But it is absolutely survivorship bias to believe that every crazy idea from a domain expert worked out.
1) That if the crazy idea is from a domain expert, it deserves to be taken seriously. It doesn't deserve to be taken as gospel, but it shouldn't be dismissed outright.
2) He's betting (but not stating as fact) that if you did blindly support the crazy ideas of domain experts, you'd be right more often than you're wrong.
This is not the same thing as saying domain experts are always right. Simply that all things being equal, you're better off betting on their expertise over your own ignorance. But even better than that is to look into the matter until you have sufficient expertise to properly evaluate their idea.
> Such ideas are not guaranteed to work. But they don't have to be. They just have to be sufficiently good bets — to have sufficiently high expected value. And I think on average they do.
It's likely that you're both right. If something is biased 45% off of a coin toss, but gives 25x returns, it's still positive EV.
This is a part of VC and speculative investment that a lot of people gloss over. Success is the difference between being wrong 95% of the time and being wrong 97% of the time. (And, with Softbank around, sometimes it's the difference between being wrong 97% of the time and 99.5% of the time.)
Who believes that?
Where the author suggest a rule: "Such ideas are not guaranteed to work. But they don't have to be. They just have to be sufficiently good bets — to have sufficiently high expected value." - it is hedged against survivorship bias.
Interesting. Do VCs outperform the S&P 500 in general?