I feel like this is one of the most advantaged times in history in terms of regular citizens having access to cutting edge tools.
Looking online it seems like the low end estimate might be $30k a year for such math researchers? And ChatGPT pro or whatever you want will run $100 a month, and should be coverable by grants. I’m quite sure matlab alone cost more in the past
Such suspicious phrasing lol. So you’re saying Paul Graham and his wife Jessica have 800 MILLION dollars worth of OpenAI stock, and that’s not so significant?
Has The Information broken any critical news about OpenAI? I never connected the dots around why I started finding it increasingly in worth paying for over the last year or two, but editorial bias feels correct.
Price targeting can help the poor in some cases and hurt them in others. For essentials where the need to purchase is high and the provider has a semi-monopoly, dynamic pricing leaves everyone worse off. For instance, think of groceries where there is only one store nearby or medicines with only one producer.
On the other hand, for something like a Netflix subscription, price discrimination DOES tend to help the poor users out. Netflix is 10x cheaper in third world countries for the exact same product. If they were forced to charge the same price everywhere, they would just charge everyone the US price and foreign users would be left out.
Per customer pricing will squeeze every customer for every dollar they can possibly afford. The more data they have the more they can calculate the level of desperation for each purchase. If they have your message history and see your mum is dying, they can spike flight tickets for example. And they will know exactly the highest amount you can afford for it.
I would say it would "squeeze" every customer for every dollar they are willing to pay. Perhaps semantics, apologies if so.
I have no problem with this for luxuries like Netflix which have sufficient competition (I am not saying Netflix has sufficient competition - I don't watch much TV, but I assume there is at least some: HBO Max, Disney+, others?)
I think I have a problem with this for literal necessities such as food, water, air.
I believe the solution to these problems is competition. If there is only one grocery store available to me, that store can set prices at whatever they want. If there are 20 stores near me - they are going to have to compete on price. I know plenty of very wealthy people near me who still get many of their groceries from Walmart.
If my mother is dying, but there are literally 1000 safe, fast ways for me to get to her in a hurry, the price is going to be reasonable unless there is some legislation which enforces some minimum (which I am also against).
That does not I mean I in any way support data collection without consent.
I do agree with you that with personal data, and without competition, for-profit companies have a strong incentive to, and will, "squeeze" you for every dollar they can get.
I think some industries, such as energy, are naturally resistant to competition; and while I am generally wary of regulations, those are areas where I think regulation in the public interest makes sense. The question of course becomes which products and services are naturally resistant to competition, and necessary enough that regulation should be required. I don't think entertainment falls in this category. I don't know enough about airline travel to give an informed opinion.
Price discrimination at all is not the same as individualized prices. And really the issue conflates two things: 1, privacy and surveillance pricing; 2, AI profit-maximizing.
Even if Netflix or others do price-discrimination, the AI-pricing issue would still be used to squeeze as much as possible from the poor. It's not like these blood-sucking capitalists who run these massive corporations are into helping the poor.
Yeah I just don't buy that it would somehow help AI companies for everyone to be existentially afraid of their technology. It seems much more reasonable to think that they really believe the things they're saying, than that it's some kind of 4d chess.
Additionally Dario has just been really accurate with his predictions so far. For instance in early 2025 he predicted that nearly 100% of code would be written with AI in 2026.
I think if you just look at what people like e.g. Sam Altman are doing it's clear that they don't believe everything that they're saying regarding AI safety.
> nearly 100% of code would be written with AI in 2026
I feel like this is kind of a meaningless metric. Or at least, it's very difficult to measure. There's a spectrum of "let AI write the code" from "don't ever even look at the code produced" to "carefully review all the output and have AI iterate on it".
Also, it seems possible as time goes on people will _stop_ using AI to write code as much, or at least shift more to the right side of that spectrum, as we start to discover all kinds of problems caused by AI-authored code with little to no human oversight.
It helps with sales because they position it as “we can give you the power to end the world.” There’s plenty of people who want to wield that sort of power. It doesn’t have to be 4D chess. Maybe they are being genuine. But it is helping sales.
Isn't it more: "We can give you the power to eliminate the people in your organization you dont like" and expands into basically dismantling all government & business for the benefit of the guy with the largest wallet?
It's hard to see as anything but a button anyone with enough money can press and suddenly replace the people that annoy them (first digitally then likely, into flesh).
They're not saying today's AI has that kind of power, and they're not saying future superintelligent AI will give you that power. They're saying it will take all power from you, and possibly end you.
If this is some kind of twisted marketing, it's unprecedented in history. Oil companies don't brag about climate change. Tobacco companies don't talk about giving people cancer. If AI companies wanted to talk about how powerful their AI will be, they could easily brag about ending cancer, curing aging, or solving climate change. They're doing a bit of that, but also warning it might get out of control and kill us all. They're getting legislators riled up about things like limiting data centers.
People saying this aren't just company CEOs. It's researchers who've been studying AI alignment for decades, writing peer reviewed papers and doing experiments. It's people like Geoffrey Hinton, who basically invented deep learning and quit his high-paying job at Google so he could talk freely about how dangerous this is.
This idea that it's a marketing stunt is a giant pile of cope, because people don't want to believe that humanity could possibly be this stupid.
Exxon has never bragged to investors that they'd burn so much oil, civilization would collapse from climate change. They've always talked about how great fossil fuels are for the economy and our living standards. It makes no sense to sell apocalypse to investors either.
There are all sorts of things you could do that might make an AI like you, and none of them have more justification than any other. This is not an argument AI firms are making.
I agree that short-term greed is driving investment, but it would drive just as much investment if AI companies were not warning of apocalypse. Probably it would drive even more, because there'd be less risk of regulatory interference, and more future profit to discount into the present.
So why are they making those warnings? It doesn't benefit them. The simplest explanation is that this stuff actually is dangerous, and people who know that are worried.
> So why are they making those warnings? It doesn't benefit them.
Because "we built a chatbot that can generate technical debt" is not a good proposition for investors. "Invest into our AI before it takes over the world and fires all knowledge workers" is.
> The simplest explanation is that this stuff actually is dangerous, and people who know that are worried.
Does anyone have good estimates of what percent of real production code is currently being written by LLMs? (& presumably this is rather different for your typical SaaS backend vs. frontend vs. device drivers vs. kernel schedulers...)
Really? In my bubble of internet news it seems the sheer number of companies that have formed and shipped LLM code to production has already surpassed existing companies. I've personally shipped dozens of (mediocre) human months or years worth of code to "production", almost certainly more than I've ever done for companies I've worked at (to be fair I've been a lot more on the SRE side for a few years now).
Depends on your reference class. There's a lot of companies and teams where it's literally 100%, and I would be surprised if there were any top company where it's below 75%. I wouldn't be terribly surprised if the industry-wide percentage were a lot lower, although I also have no idea how you'd measure that.
Perhaps it depends on what you mean by "edited by hand"? It's definitely still common for human beings to review generated code and tell Claude "no you need to do it this way". But most developers at Google, Meta, etc. no longer open up an IDE and type in code themselves.
it pushes the idea that these programs are super amazing and powerful to people who are non-technical. It also allows them to control the narrative of how exactly AI is dangerous to society. Rather than worry about the energy consumption of all these new datacenters, they can redirect attention to some far-off concern about SHODAN taking over Citadel Station and turning the inhabitants into cyber-mutants or whatever.
> nearly 100% of code would be written with AI in 2026
HN is the only place I have heard it seriously suggested that anything like this is happening or likely to happen. We certainly get a lot of cheerleading here, my guess is that in the trenches the fraction is way lower.
> Yeah I just don't buy that it would somehow help AI companies for everyone to be existentially afraid of their technology.
It makes more sense if one breaks that "everyone" into subgroups. A good first-pass split would be "investors" versus "everyone else."
From their perspective: Rich Investor Alice rushing over with bags of money because of FOMO >>> Random Person Bob suffers anxiety reading the news.
One can hone it a bit more by thinking about how it helps them gain access to politicians, media that's always willing to spread their quotes, and even just getting CEO Carol's name out there.
I'd argue if they really believed AI was an existential threat, they would shut down research and encourage everyone else to halt R&D. But then again, the Cold War happened, even over the objections of physicists like Einstein & Oppenheimer.
When your statements directly influence millions of dollars in revenue, its always 4D chess. If Sam altman beleives half the stuff he's peddling, I'd be very shocked.
> It seems much more reasonable to think that they really believe the things they're saying
It seems more reasonable to me to think that they know it's bullshit and it's just marketing. Not necessarily marketing to end users as much as investors. It's very hard to take "AGI in 3 years" seriously.
AGI in 3 years is literally not possible as it stands. Our current idea of "AI" as an LLM fundamentally will never be able to reach that goal without some absolutely massive changes
At least Dario Amodei kept the window short. When AGI fails to magically appear in 3 years he will be discredited and we can all agree that he's full of shit and treat everything he says accordingly. This is a huge improvement over the "just 10 years away" prophesying we usually get.
It's surprisingly simple to switch. I mean both products offer basically identical coding CLI experiences. Personally I've been paying for Claude max $100, and ChatGPT $20, and then just using ChatGPT to fill in the gaps. Specifically I like it for code review and when Claude is down.
I bumped from $20 -> $100 today but the Codex CLI lacking code rewind and "you can change files but ask me every time" mode from Claude Code is quite annoying. Sometimes I want to code, not vibe code lol.
I'd guess AI has made the average SWE around twice as productive at this point. This is a sort of efficiency shock, where companies suddenly need to find twice as much productive work to do or start firing employees. FB probably had a bunch of slack to absorb this but ultimately it's just hard to find that much work all at once.
I predict that tech companies will hire back a lot of this lost headcount over time. Although AI will keep getting better, so there's more downward pressure coming. Facebook, Amazon, and Google have had flat headcount since 2022, and this layoff will reduce FB's size back to 2021 levels.
I guess Meta still needs some people to run the core business (ads/social media rageslop) but your point about 2021 staffing levels would suggest they haven't been able to innovate or bring anything new to market in the past 5 years. Llama has certainly been impressive but doesn't really add more money to the pile or more eyeballs to the ad inventory.
It would be nice if someone with another big pile of money could put some of these ex-employees to work so us mid-level schlubs don't have to compete with former FOAMers (new initialism for the hyperscalers of layoffs) for 'regular' tech jobs, but it appears there are no new ideas or markets to capture.
I disagree. While their core products have stayed similar, they keep getting better at ads after Apple's privacy changes in 2021 hurt their efficiency. And Instagram has changed quite a bit, with reels growing to half of total IG usage. (Of course these are dystopian products but I'm just trying to be objective here).
To me a company at FB's scale is inevitably going to be optimizing around the margins. I mean you could argue any of Google, Amazon, FB, have had basically the same cash cows for 10+ years now.
Are you guys thinking about pushing to improve the linux software experience at all? To me that could almost be another selling point, if Framework 13 came with some downstream patches that improved sleep, power management, multi-display and hi-dpi monitor handling, etc.
And secondly how healthy is framework as a company, and to what extent do you make money from consumers vs sales to big companies?
Framework is not really a software company. But they are actively working together with various Linux distributions and they also provide funding for open source development.
It's so cool that every individual upgrade they did here can be hot-swapped back to the older designs. That's a huge extra lift that they didn't have to do.
To be specific: There's a new lower chassis, and a new chassis top with haptic touchpad. On my older framework I could buy just the chassis top to get the new touchpad. Crazy that they could make that work.
I also just really admire the CEO for doing these semi-scripted public presentations nerding out over the new devices and shouting out specific team members who did the designs. Really hope the company is doing well.
> That's a huge extra lift that they didn't have to do.
It's sorta essential imo if they want to make good on their one value-prop: repairability and the good will that comes with it. If they start releasing a tonne of SKUs with a million different parts, they'll inevitably have to sunset parts at a clip that'll completely make useless their repairability claims.
I am a happy Framework laptop owner, but I paid a premium b/c I expect moves like this. If this would change, it would become just an over-priced laptop... might as well by another Thinkpad or Dell XPS.
That said, I'm super happy they apparently have the good sense to see this. Not all companies make moves in their best interests.
Parts are open source. Other people can keep making them. There's franken think pads from China that use a x61 chassis made 15 years after the release. I had a soup of them before I moved to framework for my daily drivers.
Weird to frame that as "their one value prop" as if the quality of their offering in general is just plain better. It comes at a premium thats my only gripe with it.
My heart sank when they said 13 pro and then to see that so much is backwards compatible was amazing. It's quite refreshing to see a company live up to their mission so well.
Yeah, really impressive to see that you can take a 13 and turn it into a 13 pro with just a few new parts.
I've just ordered my own 13 pro. I've been waiting for a laptop and this ticks all the boxes. I'd previously ordered a new dell xps laptop and ultimately returned it because the keyboard was busted. I would have kept it if I could have swapped the keyboard for a new one. The use of LPCAMM is also really nice. I've hoped to see this standard start taking flight and I'm happy to grab a product with it included.
I'm not in the market for a Windows or Linux machine myself, but the way this company operates I feel like supporting them with a purchase at some point regardless.. maybe their Desktop tower
I was screaming at my screen when I heard 13 Pro but I am now ever so happy when they mentioned being able to just replace, part-by-part, a regular Framework 13 into a Framework 13 Pro.
Same, though the battery upgrade alone will be around $260 because of the new bottom cover, at that might just throw in the speaker upgrade as well for $19.
Not sure if I even want a haptic touchpad at all.
It's a matter of preferences. Actually I like trackpads that don't mind and have physical buttons. The separation between the surface that moves the pointer on screen and the surfaces that generate the clicks means that there are no misclicks and no involuntary pointer movements while clicking.
Haptic schmaptic, I just want my Framework's enormous trackpad to respect deadzones and stop detecting my palms. I had to entirely disable tap-to-click, because nothing else would work.
I might have to try their preinstalled Ubuntu images or something and see if there's some secret sauce in the input configs.
> For Linux libinput “Disable While Typing” (DWT) problems, this page claims libinput will only use the DWT setting if the keyboard and touchpad are either both identified as internal devices, or are both identified as the same device.
There is no accounting for taste.
For instance, I still prefer discrete buttons over tap-clicks or multi-finger-taps, but I would accept the mild annoyance of tap clicks over the pressing down the pad itself.
Not a huge fan of the "force touch" trackpads on newer macs, the old man yells at the clouds.
In all seriousness though I have used a pre force touch MacBook not too long ago and I prefer that experience a lot over the new one I have from work.
Though the larger size of these trackpads is something I really like and where neither the older MacBook nor the the current non-pro Framework 13 come close.
Nitpick: hot swap means without powering off. Not recommended for motherboards, batteries, RAM etc. The running electricity is the "hot" part, and without that it's just "swap".
>It's so cool that every individual upgrade they did here can be hot-swapped back to the older designs. That's a huge extra lift that they didn't have to do.
Unfortunately, as is usual for them (edit: and it makes sense; I'm not blaming them), the parts and upgrade kits aren't available for ordering (edit: or pre-ordering) yet, and likely won't be for some time, until the actual laptops are shipping. But yes, this is amazing, and the new pieces are not things I was expecting from them. As soon as it's available, I'll be taking my relatively recent AMD mainboard and putting it in a new chassis+battery+keyboard+speakers+touchpad, possibly skipping the display (I don't care much about a touchscreen, but I do care about display quality, so I'll wait for comparisons to the current 2.8k display). My laptop will, at that point, be almost entirely in a Ship of Theseus situation: I think that only the bezel and some of the expansion cards will be from the original, first-generation laptop I bought from them. That mainboard runs a number of services for me, along with an older display. A second, newer one is waiting for RAM to be a reasonable price (since the RAM it was using is now on my current mainboard); I had planned to use it for some of my research, but maybe I'll end up putting it into this older chassis and have a spare laptop again.
That all this is possible is wonderful, and a credit to them in staying true to their stated ideals.
> Unfortunately, as is usual for them, the parts and upgrade kits aren't available for ordering yet, and likely won't be for some time, until the actual laptops are shipping.
Why would you expect otherwise? I fully expect any OEM to place itself at the front of the queue for parts coming from its suppliers. If for some reason they sold parts before the laptops started shipping, I'd fully expect impatient customers would build complete machines from parts ahead of the shipping dates, which would wreak all kinds of havoc on logistics.
Yes, I really should have clarified: it makes sense, and I'm not blaming them. It's more just that, given their business model, and that they do intend to sell upgrade kits, I imagine that along with the people pre-ordering full laptops, there are quite a few of us who would be eager to pre-order upgrade kits or the parts to upgrade our current laptops.
Yes, black upgrade kit please. I have it linked and will be checking daily. For me with a 12th gen Intel, it will be worth it to have a new screen, battery, chassis and haptic touchpad. Hardware is just too expensive to upgrade, but the chassis would provide nice quality of life improvements.
Unfortunately, as is usual for them, the parts and upgrade kits aren't available for ordering yet, and likely won't be for some time, until the actual laptops are shipping.
It's unfortunate that they can't sell you something that hasn't been manufactured? That doesn't yet exist?
HN is really scraping the bottom of the barrel for things to complain about.
I had meant they aren't available even for pre-ordering, and likely won't be until the laptops are either regularly shipping rather than shipping in numbered batches, or are on a high-numbered batch. This could be months after the actual laptops start shipping. This is a process I've been through a few times at this point. It also wasn't really meant as a substantial complaint about Framework, and more just a mention of an understandable annoyance: it makes sense that they'd prioritize getting full laptops shipped.
On the other hand, nrp, since you're likely to be in this thread: if you had pre-orders and/or batched shipments of parts/upgrade kits, I would likely be paying a deposit or even the full price today, rather than ordering in a few months. Even if that meant ordering a full upgrade kit with a new display, but getting the upgrade sooner, I'd probably still go for it.
But they are available for pre-order though? At least here in NZ. I just ordered the AMD version, but the Intel one is available too (except the Ultra X9 which is already sold out).
The opposite upgrade kit: people with a mainboard they'd like to keep, but who want all the other upgrades. The kit is listed, but isn't available yet.
I have a framework 13. It looks like eventually you'll be able to upgrade the chassis to the pro one, including the battery, for under 200? Am I reading this right? That's borderline unbelievable to me
It's kind of mind boggling to me that they have a tight chassis, AND it meets their buildable/ugpradeable/repairable goals, AND their backwards compatibility is reaching back five years now.
I think a number of people would have expected these to eventually require a trade-off. Especially coming from pc-building land, where we see new non-backwards-compatible CPU and RAM sockets every 6 or so years.
There's a version of this where Frame.work said, "Design tradeoffs mean the 13 Pro is a new platform that is largely not backwards compatible, but don't worry, the 13 series will still get 5+ years of support and parts" and everyone goes "Aw, well, I guess that's reasonable."
I really want to emphasize that it's looking like Framework is creating a laptop with _better_ backwards compatibility and build-ability than a desktop PC.
All this is to say that this is very very impressive!
Even more impressive than going back 5 years is going back all the way to the first version. While I'm a software engineer, I've worked in teams where we shipped hardware, and for a consumer product with lots of constraints including implicit expectations, going against the entire trend of the past 15 years and targeting a hard-to satisfy market segment, they far exceeded what I expected when they announced their first product.
At some point they will still likely have to force that cut-off, but yeah, it's great that they seem to be able to stretch it for longer than most people would have expected.
They could have done a much more minimal version and called it a day. Being able to swap individual components of the chassis into a 5 year old model is, to me, going way above and beyond.
Back in 2002 I took the HDD from one PC, put it in a different PC, worked just fine. The worst thing that could happen is that the other one already had another disk so I had to change /etc/fstab to say "hdb" instead of "hda" and vice versa. Didn't take long for that to get fixed by specifying UUIDs and having initramfs sort it out.
IDK why it's not working for you but this should all just work without bothering with any configuration, drivers, or whatever.
This is great! Though in my case, since i have the very first generation they made, i probably need to upgrade every part of the thing so might as well just get a new one
I was really hoping for 6 USB-C ports, dual NVMe, a flush bezel, a better webcam, and most of all, I was hoping for a cooling design upgrade that doesn't cause the computer to self-roast if placed on top of a sofa due to ventilation blockage. Bleh.
Yeah completely agree. Even out of my own pocket I'd be willing to spend ~1k a month for the current AI, as compared to not having any AI at all. And I bet I could convince an employer to drop 5k a month on it for me. The consumer surplus atm is insane.
Looking online it seems like the low end estimate might be $30k a year for such math researchers? And ChatGPT pro or whatever you want will run $100 a month, and should be coverable by grants. I’m quite sure matlab alone cost more in the past
reply