I have been using Sites a few months now, really underrated feature IMO. Any time I get a quick app idea I just hand it off to Sites and usually get a working prototype within the hour.
Polished and easy to use website builders have been around for a long time, from Wordpress all the way to Wix.
The fact is, most people aren't builders and never will be. Building software is about formalizing and automating processes, and 95% of folks aren't interested in putting even the most basic effort into improving their business by approching it at a strategic level. Too focused on keeping the lights on and/or more interested in field operations.
For another example look at the many businesses using @gmail addresses to run their business while Google Workplace has streamlined custom domains for a while.
So in practice, ChatGPT Sites will mainly enable software agencies to build websites and offload maintenance to the customer : just manage it from your ChatGPT account. The equivalent of Bring your own cloud for Mom and Pop businesses, and other small companies. Not even sure ChatGPT Sites will create meaningful downward pressure on agency rates.
> Polished and easy to use website builders have been around for a long time, from Wordpress all the way to Wix.
The issue with the Wordpress and Wix is that the site stay polished for approximately 1 second after it has been initially published. In 6mo~1yr it's unmaintainable pile of dogshit that is easier to redo from scratch.
Oh come on. Sure, it might not be a production-capable system. But comparing an Astra 6/Sol 6.1 powered sites to wordpress and wix is apples and oranges.
An Astra powered site is your representation of a polished and capable business website. For the average joe, a Wix website fits the bill, their representation is not the same as yours. They focus on items such as whether or not the theme is beautiful.
I would bet there’s a (very) nontrivial number of people who have attempted to build something in Wix etc and got bogged down and frustrated enough to quit.
Lots.
I’m a pretty sophisticated user (both a designer and software engineer) and I’ve ragequit several of these tools before.
I hate Wix. I tried it to see if it could streamline things for me about a year and a half ago, and I quickly realized I could set up an Eleventy site, customize it and host it way faster. Now that agents are better and Cloudflare's API is fully exposed, I can have the agent search for a domain, buy it, set it up, set up the site and get it working on pages and probably be done before Wix onboarding was even complete.
That really depends, I needed a suspension geometry calculator, vib coded it rather quickly then did some basic layout adjustments, would have taken me a very long tone to make a anything equivalent with different tools
Really cool idea! Beat level 6! I'm still not sure how to explain how it is showing more than 3 dimensions on a 2D screen though! Would be nice to be able to choose which dimensions are represented at the various N-dimensions, but the UI might be tricky to get nice.
Yeah you have actually picked up on something that the game intentionally does to help the user: If the cheese can be reached by any of the dimensions you are currently on, then it will ensure that dimension is in the subset of dimensions you receive when you go down a dimension.
Otherwise I'd have to introduce some kind of dimension picker. Maybe a fun hardmode.
Oh I see! Thanks for explaining that - I did notice that sometimes a subset that I didn't expect was chosen.
Did the LLM explain its reasoning for how it is showing the higher dimensions? Or perhaps this is a standard way of attempting to display higher dimensional space?
I didn't ask about standard representations, but I think the graph representation works well when paired with the discrete moves this maze allows. If we wanted continuous spaces it would probably be harder to play and harder to represent, but maybe more interesting.
Same. I went to a music festival and their website was hard to use, and the schedule was difficult to understand. And the area had bad connectivity. A few prompts and I was able to build a much better offline site.
Yes, I've exported and self-hosted in the past. I'll sometimes use Sites to mockup something and then I'll download and pass it onto another agent for use in something bigger.
As a developer I almost got defensive, but I think I feel what you're feeling. Dealing with change is hard. I am a developer who once felt value for knowing these things too. Today I wouldn't even bother opening the developer pane to test CSS changes. Tools that held so much potential to me are now headed for a museum, and it sucks.
Yeah if I were to make it harder I would introduce more sparse connections requiring the user to actually understand the layout, right now you can kind of "feel" your way towards the cheese. Most brutal maze would be one with only one possible path and lots of backtracking required.
Maybe I'll do that and post to HN later, didn't expect this much response!
Got curious and don't seem to be able to reproduce what you mean? In 2D Y+ corresponds to top of the screen and Y- to the bottom. X- is left and X+ is right. In 3D and above you can rotate the shape around so there is no real up or down? But maybe I'm missing some edge case where you go up into some high dimension position then back down?
The controls for moving in the X direction are negative and positive buttons, positioned on the left and right. The controls for the Y are also left and right. this is contrary to logic because it should be on top and bottom or up and down. And the controls for the Z axis should be in opposite corners.
People seem calm in here, but I really now think is replacing a lot of industries. Like Website Designer. I totally see people asking chatgpt to build them a website instead of paying them 2,000 USD.
Soon it will connect to Ads and it will runs for it as well.
Honestly since Opus 5.5, Opus is my interface for 95% off my work. I don’t open emails anymore i let it summarize draft and handle most of the work and I only talk to Claude and not read/write the emails myself.
I also barely go in analytics dashboards anymore, Most I do is creating artifacts to explain if text is not enough for me to understand.
I’m going to bed. I have analytics running for every cancellation on my software to see what goes on. I have a/b tests running that are getting interpreted automatically. My bugs are getting fixed automatically.
I can monitor my competitors and copy every feature a/b test resolution they’ve done instantly for (3$ of my budget)
I’m an early adaptor of the latest models. The labs will have better tools already, but it really feels like a breaking point. What’s normal for me today is normal for more tomorrow and the friction has never been lower...
There was one guy commenting on my old post that stuck with me along the lines, what would you need for you to approve less. It first bothered me, but I kept it with me and I started thinking differently about my software and especially with Opus 5.5 + heavy linting and cleanup routines it seems sane to evolve to pipelinee engineering, but it also makes me feel that whatever I’m doing now is gonna be in the next model, because what I did yesterday is in the model now...
Reality is I just got turned. This is new for all of us and we all try to understand and learn this new reality. I always say what I believe most at that moment and I try a lot to not be a stuck with my old opinions, but rather have the best jugdement I can at time.
From experience my pendulum might overswing sometimes. Thank you for the reminder.
AI will also replace the "profession" of website visitor. Nobody will care about a website itself, because (1) AI will crawl it, and (2) people will just ask AI directly.
I think we have seen this sort of 'process' before, with the introduction of GPU acceleration and fast animation on UIs. It played out like this:
1. Software is fast and UI is instant, things happen instantly
2. UI is instant and software is slow - UI will lag
3. UI is animated with transitions and sofware is slow - people are initially happy, but it wears off in weeks
4. Software gets optimized, now UI animations take longer than getting the underlying data
5. The logical endpoint. Software is fast. UI needs to be instant Remove all animation and let UI render instantly; This is what most users actually prefer.
The parallel with LLM websites is direct imo. We're now getting these fancy websites, but once the magic wears off, everyone will go back to the 'text and links on page' style.
Text is not ideal in many cases. I work with CNC algorithms. I have opus 5.5 optimising algorithms. When I don't understand what it is blathering on about I ask it to sketch some svg animations to explain itself. I've been shocked at how on point most of them are. You still have to watch out for the odd hallucination but I have some base prompts that tell it always to use scripts and real math to generate SVG rather than do it by inference. With some tweaking it can create some shockingly godd one off animations to explain part of an algorithm it's working on. Text is simply difficult to comprehend when a picture can do the job much better.
I think the ability to retrieve and display data in <20ms, thereby leading to interactions that are instant is still an open and underexplored UX issue.
What if you could just hover a link and all the important data would just appear?
What if you started typing a command in console, and the execution results would appear speculatively as soon as you hit the key?
Some programs do this to some limited degree, but besides the humble textbox and outside videogames, instant interactivity is seldom seen.
That's an interesting future where everyone's personal agent basically renders their own interpretation of information that's out there. And seems like the logical end state.
People just do this anyways and we have all kinds of "bias" and "fallacy" labels to identify the ways in which they do so. We've also known about the impossibility of making people understand something in exactly the same way that you do for all of recorded history (and probably longer). Pearls before swine, can't make a horse drink, none so blind, etc. you can disseminate as much information as you want but people will ignore it, misinterpret it, reinterpret it to suit their own beliefs or potentially, even interpret it more correctly than you have (shocking, I know).
Sorry if I'm reading your comment the wrong way but it seems like this is the interpretation that many people would have (ha!), if you meant it more ambiguously then I apologize. I'd agree in the case where this is a neutral outcome where some will use it to delude themselves, many will use it in a way that doesn't stray too far from the norm and others will use it to great effect to improve their understanding.
"make the world agree and act such that I am the best at anything and never have to change, all of my current world views are what they should be" yea at that point why even connect to any other person ever. Plug straight into the narcissism machine with 5% hallucination rate...
My understanding of these predictions was more like the UI is headless and rendered from the http response/JSON etc rather than the entire thing is generated ad-hoc.
Yes. UI design is a real skill, as evidenced by the majority of UI designs being frustrating and confusing for anything but the most common use cases. AI interfaces to all these systems isn't as good as a great UI, but it's a lot easier to make and it's a lot better than a bad UI. I suspect AI interfaces are going to become the "worse is better" of the 2020s.
It's a fascinating case of use our product to build websites & apps until we can get our product to entirely replace websites & apps.
AI is this strange double-threat for folks in software development: attacking both the demand and supply at the same time.
Me, I don't think they are super eager to be a hosting platform even though this is a very logical extension for them. I think they're hoping this is a catalyst to make login with OpenAI as ubiquitous as login with Google.
Given that your entire business is what can Claude do without you doing pretty much anything, do you wonder what your market would look like if 5 competitors did the same thing?
Imagine if there were 5 competitors all selling variations of the same basic products with minor cosmetic differences! The horror! How would you ever make any money? It's simply inconceivable.
I'm not sure what you are alluding to. Is it that his margins would go down? 5, 50, 500 competitors sell the same basic products on every shopping app: Amazon, Temu, Etsy, etc. Even brick and mortar stores like Walmart sell products not of their own brand.
Or is it because he uses Claude that something changes?
Back when all clothing was hand cut and stitched out of cloth also made by hand every pair of underwear was overflowing with human touch and also cost so much time or money that you'd be lucky to have more than one pair if you weren't very wealthy. Now most clothing and especially basics like underwear and socks are mass produced and even the poorest people can afford to have new clothing on a regular basis and own multiple changes instead of wearing rough homespun and patched up rags. Yes, it became commodified. Is this a bad thing? Decidedly not unless you prefer only the wealthy to be able to afford clean, new and even stylish clothing.
Notice that commodification didn't eliminate the demand for custom made or designer clothing. It stayed a luxury and demand and variety even increased. Commodification opens up a range of markets as well besides just the cheapest market you get a whole spectrum of quality, cost and niche markets that weren't possible when supply was constrained. The same thing will happen with software. If you want to make artisanal, hand made to order apps and websites then there will be plenty of people that want that but now everyone will be able to afford their own good enough app or website without investing a huge amount of time or money to build it.
Most people host their data on cloud anyway. Once you turn off training there is no difference to storing data with OpenAI/Anthropic vs Google or Meta.
They don’t need to, and it would be too risky and too difficult to keep such a conspiracy going.
And also, what’s your story here? That they don’t delete the data they agree to delete, and stash it secretly somewhere? And that they have programs that are mire secretive tan NSA’s inside that secretly finetunes algorithms on such data?
They would expose themselves to lawsuits that would crush their companies, even with such absurd valuations? And possibly risking jail time of the top people? GDPR offences in certain countries are punishable by up to two years in prison. And with a such blatant violation at least in EU the maximum sentences would most likely be given.
I don’t mind (within reason), and I think you’ll be shocked by how many people do not care. I self host passwords and keep financial access secure. Otherwise I don’t give a shit if someone sees my 30 year old dong in a picture, my bills, or my bucket list. Have fun!
These people are beyond hope. No they don't care about anything. They are waiting for the bots to plug them into the matrix so they can finish out their lives in silico.
The killer app behind Azure and GCP in a lot of cases, the thing that really drove multi-billion dollar decisions made by the largest cloud customers in the world, was that they were not Amazon.
Why would Walmart want to let Amazon hold their business hostage, tell them all their secrets - literally pay to hand it over to them - and generate profits that would surely be re-invested in competing against them in retail? Even if AWS handled Walmart's account in 100% good faith and helped Walmart tremendously, they would still ultimately be paying to make themselves less competitive in a world where their single competitive edge (low-margin retail) would be eroded.
Would the consumer - or really anybody that likes having cheap, widely available products they don't personally make for themselves - ultimately benefit in a world where Walmart's only choice is AWS (startup-branding) and AWS(ea-branding) and AWS(IBM branding) and AWS(Microsoft-branding, if they were also evil chinese communists who maybe trained on the traces they paid for)?
It can already connect to ads. And it can look through all your competitors ads at warp speed, en masse, and figure out how well they are doing and steal their ideas and copy all their creatives.
Also, just inside Google Ads itself they stuck Gemini which tries to get you to spend your life savings on ramping up your campaigns. If you use Gemini externally it tells you Google is trying to scam you :p
Yes that’s what I’m doing. Practically i have my competitors on lock. Most of the time security lacks and you can even figure out the model/prompts/database they use so the moat is gone.
> [...] I only talk to Claude and not read/write the emails myself.
Agree on many other points, but personally, I find receiving completely AI-written emails somewhat offensive. (Personal emails, at least; for advertising slop or adjacent things I find them about equal to baseline, which is a low bar though.)
Have you received any feedback (positive or negative) on this?
I suppose once the majority of recipients also start sending every received email through an agent summarizer, it's a moot point anyway...
> Opus is my interface for 95% off my work. I don’t open emails anymore i let it summarize draft and handle most of the work and I only talk to Claude and not read/write the emails myself.
I’m trying to do the same. Could you give me some tips on what works for you?
I have the lifestyle I want since I’m 25. I have nobody to tell me what to do and can work when I want on the things I want. I don’t really have to work actually. I just like to work
They are desperate to find a moat cause models aren’t much more than a commodity.
What they might not realise is that stuff/services like these (ChatGPT Sites) are probably not much of a moat either in a world where majority of population has access to increasingly better models capable of building competitors in days if not hours. Even if they restrict usage of their models when it comes to creating competitor services there will always be other models that you can use to build competitor.
What is happening is that their products are putting people out of work but at the same time they are also enabling anyone to easily build competitors to their own products and are thus putting their own services “out of work”. They are essentially eating their own tail. Ouroboros.
Is noone seeing that?
The only truly lasting moat might be hardware/compute. At least for now.
Unfortunately right now most AI models get into the weeds of asking the user to sign up to netlify or firebase or buying a domain etc - none of which most users are interested in doing.
Look at the "beneath the surface" example. Click on a creature close-up. Click the "rotate creature" button. The creature's rectangular jpeg, complete with black background, rotates on screen. No 3d or anything. This is their demo! It has that Potemkin village quality, much like AI apps, code, and answers.
You're ridiculing and laughing at something that probably took 10 minutes to prompt and make while the same thing would take you 6 months to do. You're being ridiculous with your analysist.
This would take me 0 seconds to do because I would never do such a dumb thing. The "show anatomy" and "viewing lights" buttons would also be hilarious, if it was in a movie about a hypothetical brain dead humanity that is, and not something a big company made in real life that gets other real people angry when you mock it.
I wonder if they'll combine this with the currently-in-closed-beta Sign In with ChatGPT to let people make sites that can make inference calls that get charged to the user's ChatGPT subscription rather than the site owner's API billing account, like Claude Artifacts already lets you do. That would synergize well with ChatGPT Sites being generally more robust and feature-complete than Claude Artifacts.
They’ve both been working on this for a while, I’m surprised they still haven’t launched.
Domains, auth, and databases are probably the stickiest software products of all time. I suspect it’s a combo of security concerns and the risk of eroding goodwill among developers (their most important customer base rn) if they move up the stack too quickly. Once you know that the goal is to become cloud vendors, train on/compete with their own customers, and own the entire software stack e2e it’s hard to really feel grateful that they’re dragging it out but planning on doing it anyway.
For those of us working in infra/SaaS outside these companies it’s pretty clear that the only viable path that doesn’t involve getting cannibalized is training your own agents/models. The new coding agent-infra-data business model is a path towards full commoditization and undifferentiated prompting in 3-4 vertically integrated walled gardens. If you ever start making real money on pure software infra they’ll just be able to eat you alive by implementing something similar, training on their own tools rather than yours, and integrating it into their stack (which all of your customers are on already).
Also, you should only ever use $20-200/mo subscriptions on work you want them to train against or you don’t mind automate yourself out of. Think a little bit about what you’re teaching them to do when you use their products, especially if they’re the primary interface you’re working in. People are going to start caring about this a lot when the AI companies feel safe enough to begin the “extinguish” phase of AI coding. Train your own models!
It’s really only an existential problem for pure-software businesses/developers.
There’s a good chance that the next step after open weight models is something that makes it easier to post-train existing models on new workloads without running an entire pipeline just for yourself. Somebody has to actually build it though.
> I do not see enough people discussing this and trying to position their businesses in a way to avoid destruction at the hands of the labs.
What can most even really do about it? Assuming they pull off what they're saying they will: it won't matter what you do, what vertical, what moat you think you have. It will be a concentration of capital and power we've never seen before, and frankly that's terrible for the world.
We need to build collaborative post-training pipelines to augment models with new custom capabilities without catastrophic forgetting/excessive loss of capability in other domains.
Currently what makes this too difficult for anybody but frontier labs is the lack of access to the full distribution of workloads/data they use to RL multiple separate envs/evals without regressing more than they advance in general capabilities. Because they are the primary buyers/builders of that stuff, they have no incentive or reason to allow anybody to replay or resample it but themselves. And it's all so very expensive to do in aggregate so there's not really demand for other product shapes except more openly available, collaborative/bundled RL that you could plug your specific workloads into (who the frontier labs would obviously not want to support with their business).
However, if someone were to build a collaborative rollout platform that you could use to train private workloads (ie create useful IP that doesn't just become profit for other labs/come from what they already can do), the upfront investment would be amortized over the very large number of potential buyers once it gets into the 5-8 figure range, who essentially have no other choice if they want to remain competitive in the technology industry.
Until openai/anthropic ipo the concentration of capital/spending and ndas/loss of employability is too concentrated for the best researchers to really do this without rocking the boat. And a lot of also-ran ai/saas have the same risk due to the lack of capitalized acquisition opportunities or AI vendors to partner with.
If the plan is ultimately to drive you out of business if you ever build anything profitable with their products, and monetize your knowledge without fairly rewarding or explaining their intent to do so, you might as well defect early and build what you inevitably would need anyway.
I think that overlooks requirements like "absolutely ensure the end-user has to have seen and agreed to this exact text before billing them, because otherwise it's a massive fine."
Pessimistically, companies will ignore that and pick the cheapest solution and accept an unknown failure-rate that they assume will be tiny.
Optimistically, they will demand "the AI you just demo'ed, except fix whatever bug so that it never screws up." In that case, SaaS will not be about "training your own LLM model", but about investing in harnesses and rule-engines and GOFAI that constrain the actions of a not-necessarily-custom LLM which is acting as the consumer's UI or concierge. In other words.
On one hand, I am incredibly tired of seeing the absolute train wreck of waste that is suave founders taking millions and blowing it on hundreds of personnel who do practically nothing— especially having seen what a single person can do with these tools.
On the other hand, death of a livelihood in a social structure that doesn’t care about you.
We’ve always had pump-and-dump trash. The terrible stuff loses and the good stuff wins. I don’t see why that would change.
The companies that ditch the management structures of yore and successfully build better software faster with the new tools will come out on top. I don’t think these companies will replace ALL saas subscriptions with in-house solutions, because one huge benefit of paying someone else to hold that bag is liability and regulatory overhead.
Honest question for devs out there... is anyone worried? I don't see enough frank conversation happening around this subject. Timelines. How long until salaries fall? Is it too early to stalk thinking about back up plans? Am I crazy?
No, you should be thinking about backup plans, because you should assume you’re not going to walk into another career (I am not suggesting that you are).
- People in other industries actually have specialist knowledge (this is something IT people forget)
- AI is coming for those jobs too
- and so are lots of other people, not just displaced techies but from any number of other industries.
Thinking about it in advance should include training yourself up for other sectors.
If AI is successful, it will be a very different world of jobhunting. If it’s unsuccessful there will still be job losses for a while because of overspending, and because of the wider economic issues.
It is not clear to me that it is yet at a level of success to cause massive tech job losses (it's just causing large-scale misery).
It's also not clear to me that the level of success in the tech industry is truly being repeated anywhere else. There are industries where generative AI is actually a liability (the law, finance) and where it is increasingly viewed as repugnant (design, the arts, to some extent photography where deep learning noise reduction tools are OK but the culture is swinging hard against other generative uses).
I am also not convinced it is "objectively" successful in the tech industry, considering the extent to which most token pricing is subsidised and people's reluctance to demonstrate ROI even on subsidised/bundled tokens.
I'm not claiming it is unsuccessful. Just that I don't think the job apocalypse is here yet. Though this product (ChatGPT Sites) is very indicative of a coming tech job apocalypse in pure software services.
It's certainly at a level of success where people should start thinking about a "Go bag" on an employment level. Keep your CV up to date, work on gaps in non-tech skills, consider other qualifications.
Coding is probably solved. Honestly it’s probably going to be 5 years of runtime starting now depending on industry lag and uptake.
A team I was managing spent a whole year converting government forms into a web app. It took a while and there was a lot of discussion with the end users and a bunch of back and forth (keep in mind this is government software with a lot of onerous legal requirements). Then about 4 months ago, we just fed Codex the forms, along with the UI design and the legal requirements and it spat it out in 1 hour. Was it perfect? No. But our iterative loop became make the web app, show it to end users and then incorporate feedback and completely cut out the UX team, product etc. We were able to complete the conversion in 2 weeks with one guy (something like 150+ government forms/applications). For most simple to medium corporate development work it’s basically over.
We're in a similar boat. Entire eng team went full AI about a year ago, and honestly we're struggling to find things to do. Features that would take weeks, take minutes.
I just don't see how we keep going like this without layoffs. The company cannot responsibly continue paying for this many engineers, when half, or less, could do the same job.
Serious question from a consumer of software: if software can be improved so effortlessly now, why isn't it getting better?
Like, why is it all still so buggy? Why does no website seem to consider 96 PPI displays in its design? Why don't we have ostensibly useful features like undo trees everywhere? Why is everything still a RAM hog? Why does the proportion of websites that pay no heed to accessibility seem to be going up rather than down?
I'm told we're living in the future, but it sure doesn't feel like it.
For the same reason that increasing headcount never really solved the bug problem.
Flawless software isn't the goal, ever. Feature development is almost always the primary goal, and fixing bugs and performance issues is only done if the bug is bad enough.
I don't think you're wrong overall, and it jives with tripleee's reply, but I don't think wanting end-user software to be less buggy than it currently is is asking for "flawless" software.
You're right, I was being hyperbolic. Bugginess in software is a spectrum, and asking why we're not further along that spectrum toward being bug free, is really what the OP is asking. My answer still applies.
I live in the future when it comes to my hobby of music making. I make my own little software tools that I did not have the ability to make 2 years ago. Little tools that are too specialized and specific for my obscure taste to have much wider economic value as tools in themselves.
Claude is also the greatest obscure book recommendation engine I have ever found so I have a lot more interesting books to read than in the past.
Outside of this, I basically live in 2019 but everything is more expensive.
I worry this is the future. It doesn't really feel much different other than things getting more and more expensive.
Lack of information. When you are on the building side you become blind to its faults. Consumers almost never tell you what isn't good about it.
> why is it all still so buggy?
Also lack of information. I haven't seen any bugs in years in the things I've worked on. LLMs have closed every last obvious testing gap, squashing any outstanding bugs there. That's not to say it is all bug-free now, but what bugs users may be experiencing aren't being reported.
AI is good at building software, but it is still not good at consuming software like actual consumers do.
Because the goal isn't to make the future better for humanity. It's to make the future more addictive for humanity. Big difference. More clicks, more ad views, more engagement, not necessarily more pleasant. I mean, that's obvious.
If your company isn't throwing you into R&D right now everyone is looking at layoffs. You are no longer in an environment where doing "the same job" will keep you afloat.
I have been the lone developer on a project for quite some time and was already feeling that way. Yet, recently found out that a second is being added to the team and a third is to come. I guess I'll enjoy what is effectively a paid vacation before it all comes crumbling down.
You're not crazy. I think there's still room for the top 5% of devs working on genuinely hard technical problems.. but I don't see a bright future for the rest.
Salaries won't fall, they'll go up. The number of roles is what'll fall. SWE is going to become a giant game of musical chairs, and people who can't find a seat will get pushed into human facing roles ("product" / "technical sales" / "customer support" ).
It'd be hard to compete when there are too many sellers like you. In many businesses, data is the moat and network effects keep making them more valuable over time.
There’s going to be a lot of chaos at some point but I think that “hybrid builders” (scientists, sales people, product developers, etc). true specialists, and new opportunities in hw/robotics/training/ will soak up a lot of talent, and be rewarding to work in. Feels very sadly k-shaped.
Software engineering on one end is going to be more like finance or medicine where the cost of poor performance is actually higher than the risk that they wont be worth what you pay them (ie there isn’t any amount of pay where it’s worth it to hire them). Then on the other it’ll be like knowing how to use PowerPoint. All the new tech categories have too many bottlenecks and challenges beyond just producing code to consolidate like that.
I think technically building this stuff is now easier for the masses. But knowing what to build, how to describe it, how to test it, figuring out how to understand the users -- that's the hard part, always was and I'll be employed doing that for a couple more years at least. But after that, I see this all going towards building for agents, building for users will be a thing of the past. Then I think its not as easy for people to relate to that, probably want a machine designing for other machines. UX and taste don't really matter much at that point.
No. Saas hedge is to offer solid api + mcp for agents to use whilst still maintaining their own products alongside. That's it, that's the whole thing.
Are you going to trust a vendor who has been running for years with your business critical operations or brad_69's awesome saas skills and agents? Do you think these businesses are going to shift their whole product to be in Altman's walled garden when they don't need to?
I've said it before but I'll repeat it here until someone changes my mind: jobs in IT are safe until using AI gets its equivalent of "no one ever got fired for buying IBM". Until there can be financial and legal responsibility for when things go bad, people are going to be needed.
Once it's so good AI labs will be happy to accept liability, it's game over. It was never about code quality.
Slapping up a slop website is no threat to professional tech folk.
When they have full DevOps, with multiple environments for testing, compliance with security best practices, can do database updates and data migrations, move the app smoothly between different platforms, streamline costs, and scale to at least hundreds of thousands (if not millions) of users without crashing and burning... then go ahead and worry. We aren't anywhere near that.
On the other hand, if you are purely a coder, and cannot run a project that does all of the above... maybe start your worrying earlier.
We aren't, but the AI Labs no doubt are. They just can't release it right now because their bread and butter is still developers burning tokens, so there is still a business problem to solve.
I think this has always been the end result - when the user can enter a business problem and it spins up a site, and they don't have to think about the hosting, database, app, etc.
> And it will be good enough for 90% of use cases.
I think it’ll just raise the automation bar that much higher. In other words, it will be good enough for 90% of yesterday’s use-cases. But tomorrow’s? I’m bearish.
I imagine the amount of onerous bureaucracy when interacting with any entity is going to explode if they expect you to do it in an automated fashion, for both private companies and the government. Especially as roles are cut because they can be “automated”. Except only 80% of the workload can be reliably automated, the other 20% requires a long tail of effort.
Its not for those uses cases. though. THink about squarespace/wix hosting all the company landing pages, restaurant menus, daycares. This will replace it.
It’s mainly mucus, from a piece of mouse colon that was cut open and gently flushed out, leaving the inner mucus layer attached. The actual mucus is black here by default but there is another channel you can turn on in the sidebar to see it.
On top of it we added bacteria (green) and bacteria sized beads (red). On the bottom are colon lining cells (blue).
The inner mucus layer of the colon was generally thought to be impenetrable to bacteria, serving as a way to protect you from the trillions of bacteria in the gut.
It turns out that actually, it’s really easy for bacteria to swim through mucus by using flagella. We found some bacterial mutants that swim through mucus (the green ones). The red beads serve as a control marking the top of the mucus since they can’t get through the mucus.
The real question is then - if they can get through the mucus at any time, why don’t they? And there are various theories about this - which I’m happy to elaborate on this you want. The basic idea is there is a risk whenever you cross the mucus (you get closer to the immune system) and mucus actually feeds and tames bacteria over time by providing them with various sugars and things like that. But if they get desperate/starve, then they will cross.
Mucus is still a mystery - there are substructures within mucus, take a look at this image for example:
I don't think this provides a lot of value for more complex applications. It just lowers the bar for getting started.
Once the app gets a bit more complex, it won't be that much different to any more traditionally vibe coded application (code, git/pr, test, deploy, ops). It only removes the barrier to get started, but doesn't solve any of the real problems. Problems like AI slop, code rot, bad architecture, security issues, data loss/breach because of bad code, and so on.
That's my experience as well. Some folks like to dabble, but for the most part they have other stuff they want to focus on, they don't want to spend the time to ensure it's good and it works.
I'm not saying an agent couldn't make sure it's good and it works, but there are a lot of aspects to that: It's compliant, it's secure, the right data is stored, the right compromises are made for sometimes mutually exclusive business priorities, UX is decent, it's responsive, it scales, hosting is economical, it has the right tests, monitoring for the right risks, and so on and so forth. A lot of devil in the details, and the context required to know what to focus on is rather vast.
If anything, I see this work getting harder, since the discipline around minimal scope is getting weaker, leading to a complexity explosion affecting all of these aspects.
It usually comes down to how successful a business is. Just starting out with little to no funding? They'll want to dabble themselves. But if there's money, it can and will be exchanged for time. Easier prototypes and MVPs means faster business model validation, which means higher willingness to invest. I do see this happening.
Will we always need to know programming languages for doing this? Maybe not, but it doesn't matter. If you make a machine do stuff for a living, you're a programmer. How exactly you do it doesn't affect that in my mind.
The Saaspocalypse was a wet fart that never went anywhere. People aren't using these AI tools to kill or replace SaaS at their jobs as much as HN wants us to believe.
The argument there wasn't that everyone everywhere would use individually customized software* instead of packaged SaaS; it was that, because AI greatly reduces the cost of building an internal app or service, it improves enterprises' BATNA when dealing with SaaS vendors, which means surplus will shift from producers to consumers. It's not obvious to me that that's not happening, since it may have been masked by a general increase in automation since AI also synergizes significantly with SaaS. (The Stripe Economics post gets into this a little, though it's mostly less curious than I'd like about why the "SaaSpocalypse" didn't materialize. Of course, Stripe has an economic interest here.)
* Some people do think this (e.g., https://sockpuppet.org/blog/2026/09/25/what-even-is-an-os-no...), but I remain somewhat skeptical that this'll be true for the most economically significant applications. This is not because I'm an AI capabilities skeptic (I'm not, I'm more of a doomer), but because even in a world where AI is designing and implementing all the software, there are probably still significant benefits to standardization. I see it as sort of analogous to compilers; we might get a world where compilers are both used and maintained almost exclusively by LLMs, but we probably won't get one where compilers stop being used or maintained at all because LLMs write everything in assembly.
In stuff like wikis and ticket systems what I see is people now just making their own little tools around the ones we already use, this arguably just sements their position even deeper.
Also I dont get the inhouse argument, you dont really save anything. So instead of paying X to SaaS, you now propose to pay X to anthropic instead AND maintain said SaaS... what? how is that better, that just shifted the cost to another company and you got the burden of maintenance and responsibility.
I mean, with a Cloudflare API key you can pretty much already do that. But keeping your data straight and putting in the right authentication guards, etc. are all still where the business problems meet technical and the AI is not enough unless the agent is both architect and customer.
That's what I do but normal people don't know what any of those terms mean so their daily AI app does that shopping for them and they don't need to get an "Appy Key".
People curate their brands on socials. Having a website is not really the badge it once was, and the ability to generate them like this further reduces the social proof they used to provide.
Look at squarespace + wix + lovable + framer + vercel + wordpress (and agencies), cloudflare, godaddy, etc. and think about how much more demand there is for a vertically integrated one stop shop without friction. This is a massive market even if it doesn’t become myspace.
This is the top of the funnel for most people building a tech product or marketing their business, so it’s very helpful if you intend to sell other services they might want to use down the line. (Cf Cloudflare except imagine if they also integrated ads into one of the most popular websites in the world)
You want to talk to a few guys not in the software industry to understand what the world is like first.
While there are many enthusiasts out there, lots and lots of people couldn't care less about ChatGPT allows them to "buuld their own website". It's simply irrelevant to their life.
Gonna wager good money they attempt to relaunch Apps or something of the sort, where you build, launch, host within ChatGpt, seamlessly making it easier really soon.
I think there’s a bunch of startups that were basically this. This is the problem with AI startups, you’re just waiting until the labs will come out with what you were building and make you obsolete.
I don’t think it’s all bad though. The labs aren’t going to release anything that makes sense if you’re trying to be vendor independent. So, gateways, vendor agnostic sandboxes, harnesses, etc. That’s where business opportunities may be.
I don’t think people understand the existential threat this poses towards Microsoft Office, Gsuit and Conflunce.
We have internal service like this at $day_job which has become the default “information” authoring and sharing platform for almost all staff.
Information is more likely to be shared on this tool than through Confluence or an MS Word/PP artifact because HTML docs are so expressive and all friction in writimg HTML documents has disappeared.
The idea is great but the websites shown as examples are clearly made by AI because they make no sense.
Current AI is NOT going to replace web devs or coders because it cannot tabula rasa make human sense. I suspect the reverse with a Jevons thingy happening
Idk about that, a company like Form3, Edenred etc are technically just software companies - but no one in their right mind is going to build a black box payment provider. Sure, if your business is wordpress websites, you're probably going to see a big downswing in new customers. But there is absolutely still going to be a place for highly specialised and security dependent software.
something that solves a real problem for businesses, that you are uniquely positioned to execute on well.
If you target the lowest common denominator, or consumers, or build something really quickly - good luck
For whatever reason most of the sites are Arabic-language prototypes for things like plumbing companies. I'm not sure why - if those are solving a real problem, or whether someone is running a scam, or what.
Cool project, but the hero art is a little strange (what is the dude on the left hammering? Aren’t the ones on the right both pulling the pulley down from opposite ends?)
Thanks! I think the fox is steadying the log as the raccoon hoists it. Agree the badger is sort of hammering the air, haha. The art was AI-generated but with some manual editing in Affinity, so hopefully it's not too sloppy.
Claude Artifacts and ChatGPT sites are one of the most useful features. The models are great on writing HTML reports.
However, I change the models too often, so I didn't want to be locked in a specific platform.
That is why I built getonup. It is a way to quickly deploy static pages using Cloudflare Workers on your own domain.
Just point your agent to the getonup repository and ask it to setup on your Cloudflare account on a certain domain (like pages.vitorbaptista.com). It will configure it and setup Cloudflare Zero Trust to secure your site. Then you'll have a date-ordered index (so you can see every artifact you deployed), and you can deploy static sites.
It is still very simple, but it has already been extremely useful for myself my team. We often publish multiple pages per day.
I’ve been using exe.dev to throw up micro sites. It has a built in web agent and first party mobile app agent. And any coding agent that can use “ssh” (all of them) can easily figure out how to drive it.
It also has a lot more developer and PaaS tools like GitHub integration.
RIP Heroku for imploding before the wave of vibe coding. We had the ability to “spin up a free app on the internet” before it was cool.
Very polished. I suspect OpenAI will abandon this project at some point to focus on other core products. This kind of feature is great, but it spreads them thin, and other players like Cloudflare are rolling hard to provide this kind of convenience to agents with far greater depth. This is not a judgement against ChatGPT Sites - it's a well-executed product.
Sites has been one of the most valuable ships for us internally outside of general model releases imo. A few things that I think are incredibly underrated:
1. Plugin usage. I can build a site with some data vis and when someone else in the company logs in, it uses their access to evaluate those queries. The access model/BYO connection is really powerful.
2. Ephemeral sites. Setting up a site is so easy they can be ephemeral. I can set up a whole site for a presentation, or a complex form, or just a meme, and then never touch it again because it cost me basically none of my time.
3. Sites as plugins. Sites directly integrate with the plugin ecosystem in that a site itself can provide/be a plugin. Combine that with Sign in with Chat (byo subscription) and you can build some really cool experiences really easily. I've been working on a deck grader for Magic the Gathering the last few months for my own uses and now I could pretty trivially publish it as a site for others to use.
I don't see sites being abandoned any time soon :)
I took it for a spin. I used Claude to create a self contained HTML file with JSON data embedded to create a SVG bar chart with sort and filter buttons. I then gave the HTML file to ChatGPT Work and it created the site using the provided HTML code unadulterated.
I then asked the next question. How can I automatically update the embedded JSON data. It gave two answers which I thought was weird. First, it said I could using a plugin to an MCP server. Second, it said I could use an API and when I asked for which API... it said there was an API available in ChatGPT Sites but no details. So I assume it can be done.
What I am thinking of is to use ChatGPT sites as a replacement for SharePoint. I can create self contained HTML charts and let employees view it within our enterprise instance of ChatGPT. The piece I am working on is how to automatically update the embedded data from our database.
It should have suggested using Scheduled Tasks, a feature in ChatGPT to run a prompt on a regular interval. The task can fetch up to date data, then make changes to the site and redeploy. Alternatively, if you make the site full-stack, and use the provided DB, the task can update your site by writing to the db directly or by calling an API endpoint on your site.
Something I've been thinking a lot about recently is how much we still need UIs vs if we'll just interact with agents.
We have a platform that enables building both and I still find UIs to be quite relevant. Especially if it's so easy to build like in this case because you can always throw together a visualization for something quickly and then never look at it again.
This looks really cool and because I'm in a tangential space I know how fun and useful it's been to build and host stuff on the web quickly.
We're geared towards fully-fledged (serverless) apps and agents for internal use only though, in case it's of interest to anyone: https://railcode.dev/
I was imagining the other day what other "apps" that frontier labs can get into. If we had "Claude Mail" or "ChatGPT for Workspaces" it would be very interesting. Email service created by Claude, then docs (they have it already?) then all the business apps layer (heck integrate your vibe coded apps onto these! build your own slack, CRM)
If models are commoditized and apps is where the grab is - I actually wouldn't mind a new email provider - GMail has not practically changed in the last 10 years I feel.
The big picture here is that LLMs using APIs to get into information stores works but is stupidly slow. Email is a good example. You can hookup gmail or whatever to codex or claude and it's great. But it takes ages for it to dig through your inbox and that involves doing a lot of API calls. Most SAAS tools are about gate keeping data via APIs. You need to break open those silos to get to the data. A CRM is just a contact book with a lot of meta data. Why not download all that data and provide direct access? It's a lot faster than accessing records one by one via some API.
What if email existed on your laptop? Like was common before we moved to using web based applications? What if your local email client had a local database with a proper search API. That sounds a lot like what Thunderbird, Outlook, etc. did a quarter century ago. And still do if you still use those. Digging through a local database is a lot faster than dealing with relatively slow API access. LLMs are really good at doing stuff like that.
I run a small company, I've taken a sledge hammer to many of the SAAS tools we use in our business in the last years. We got rid of Hubspot, Slack, and many other tools we use for financial planning, etc. These got replaced with self hosted open source things or simple file based repositories in some cases. This has several huge advantages: 1) it's stupidly easy to get to the data for LLMs now 2) LLMs can administer these things for us 3) we saved quite a bit in licensing fees. 4) removing obstacles means LLMs can do complex things we used to do manually in dreadful SAAS tools.
I'm not sure a chat interface is a good UX for a lot of use cases. I want my emails as a list of emails, not an explanation of what that list contains
I saw they're rolling out the done-to-death personal finance Mint-competitor and it's the exact same issue. At first it was interesting.. then when it comes to actually using it it's just not a fluid UX
I think the reason there isn’t much panic over this among web designers is that we are in the long tail of web development.
Most websites are basically pointless, more like pretty “brochures”.
In the near future, useful structured information will be dumped to websites that are funded through micropayments using x402 protocol. Agents will be the ones making these purchase decisions, replacing revenue historically brought by ads. Will be interesting to see the death of programmatic advertising on the web.
I just had a tangential thought, but what if... AI completely recreated their own version of the internet from scratch? And that's what the internet eventually becomes.
My question ("Do you know what <details> and <summary> tags are?") doesn't appear to be here; but given the page functionality the answer seems to be no.
I get that this can be considered a "minor inconvenience", but I think it's pretty embarrassing on a page that's supposed to be marketing a website builder that's supposed to take care of knowing these things for you.
I think this is the next step in the evolution of AI based tools. We can develop the app or tool, but we need somewhere to easily publish it.
I actually recently shared my own version of this concept, but self hosted. I'm optimistic that there can't be much lock-in from the big providers, but I think there's a large market to capture
I've been having such a blast making simple mapping apps that aggregate public data like GIS sources. I made a map for my city with every public parking space labeled, interactive with links to google maps and free hours, et cetera. I also made an educational map for myself to explore data on Jihadism.
Sites is so good that it’s changed my workflow - I start in Opus 5.5, have it write a plan to Google Docs in markdown, then have Codex pick up the plan, shave it down, and build the prototype. A little annoying to switch tools, annoying enough to only do this on desktop, but the output is fantastic.
Why not just… have Opus 5.5 build you the website? I would wager that there is minimal value add by passing things back and forth to different services vs just refining in one. But that’s just my personal opinion/experience.
Tbh I just fed a couple paragraphs of "big picture" specs to ChatGPT, and asked it to give me a solid design doc / spec sheet, and serve it as a markdown file.
Then I fed it to sites, and out came a working product.
Depending on the idea this would be a great idea. Simple sites like portfolios for example are perfect for this job. Most people however are usually vague when describing what they want so I imagine you still need to be able to structure your toughts correctly. I'll definetly be trying this
We’ve been building something like this, but perhaps even more powerful, safe, and beginner friendly. If Sites looks like it’s the right speed for you, we’d love your feedback on https://vibes.diy/
They'd better be proactive with checking for abusive users, because I can easily see people trying to spin up honeypots, crypto miners, scam sites, etc. on this. Especially given that the hosting appears to be free with undefined CPU/RAM limits.
I have been building https://github.com/openrundev/openrun. OpenRun allows you to build web apps, adding OIDC/SAML with RBAC to control who can access the apps. Supports MCP for app management. Also supports service bindings, which enables providing each app an isolated database account.
My guess is Google and Apple are going to eventually take all the market share for this "app builder" category. They own the device market-share and can offer way better integration
You think thanks to AI companies are going to be soon building laptops that compete with Macbook Airs on build quality, performance, battery efficiency and compete on price? Or tablets that compete similarly with iPads?
I mean, fair opinion to hold. I don't think AI solves most of the things that were preventing Samsung and HP from doing this so far. Which ones do you believe it solves?
Does anyone know if Claude makes it easy to publish sites u vibe coded to a custom subdomain or new domain? That’s the biggest reason I use this feature in ChatGPT over Claude
You can ask Claude to deploy it with ShipStatic. It works pretty well for this use case, no account is required, and you can publish to a custom domain/subdomain without much setup.
Disclaimer: I’m one of the people behind ShipStatic.
[Shameless plug]
The problem with current llm based website builders is that they all end up looking the same, almost have the same flow and structure.
It's worth for the token budget. You can do so much with gpt-6.1-sol (high) on a small sub. If the money is not aan issue it's probably still worth it for Astra because it can occasionally be smarter than Fable 5.1 (or so I heard).
I'm genuinely sad that what we learned with hard work and sweat, our craft, our identity, our soul; is now something you can buy. I've been in this profession for a long time and now I see people who have absolutely no fucking clue, but lots of money, and they buy all these agents and make things that work (way more broken, with absolutely shit quality), but they work and so they brag about their creations to me all day long, and man that makes me feel like a total loser.
I spent decades trying to master media theory, language design, computer graphics, compiler design, rewriting systems, mathematics, logic, build systems, graphs, data-flow systems, making GUI kits, layout systems, ... all to make a better system for thinking, and now the whole idea of thinking is outsourced (as altman says "sold as utility by the meter"). I need to become an addict or something and be wasted. This is not a life.
If it's any consolation, almost no craft will be spared. Software was simply first in line because of how much training data was available, but it won't be the last.
Are we looking at the same website? My eyes immediately go to the font and typeface styling and it registers immediately as AI generated. There’s a very distinct “AI” website aura that Claude and ChatGPT make that requires you to change a ton of the styling to make go away.
I think people are missing the point. To fix the sloppiness all you need is a non-technical person who wanted the site/app to tell it what to fix. Either way the technical middle-person is gone
You spent decades on this and you think that linked website is not slop? You say its "basic" but its really not, its overengineered nonsense if you look closely.
Could you just be an API to serve intelligence instead of moving into the app layer and putting your clients out of business. How much wealth and power do you need to accumulate seriously?
The "intelligence API" layer seems like a pretty rough business to be in. The smarter the models the more they're commodities (since they can use any harness or scaffolding) resulting in a race to the bottom on pricing. And then each new better model requires double the cost of the previous one.
I'm surprised there's not a bigger push into better and useful memory, since that would result in actual lock-in.
What is wrong with being a commodity, or a utility? Oil companies do just fine. So do electricity companies. Why does everyone have to seek world domination instead of building a good business. My point is by cannibalizing their clients they risk not being trusted by anyone. But Microsoft did the same thing and created a durable monopoly because Republicans abandoned anti trust enforcement and that seems likely here.
The question of them moving into the app layer is completely irrelevant. If you are threatened by this the reality you need to begin making peace with is that your business is not going to exist within 10 years. And that's being extremely generous.
Irrelevant to what? It’s highly relevant to how AI will evolve and how the distribution of its benefits will propagate across society.
If they merely acted as APIs, even if that meant they would be “relegated” to the status of a utility then individuals and businesses would have faith and confidence that their use of AI wouldn’t be the very thing that puts them out of business. I think that would result in more innovation, faster adoption and deployment, and broader distribution of economic benefits.
As OpenAI and anthropic march towards an everything app they are smothering infant industries in the cradle and undermining their own users confidence that it’s safe to use their models.
I have talked to over 10 people at universities, SMBs, and large enterprises that are actively allocating millions of dollars to bring AI in house.
No it won't matter because if they don't someone else will.
It's like trying to stop adoption of computers by cobbling together a union of human calculators. It's an economically nonsensical position that simply comes across as denialism. It's not even desirable.
You are advocating a protection of the managerial and owner class to keep the economy needlessly balkanized. Universities are some of the greatest offenders in perpetuating inequality. Why have universities gotten more unaffordable?
Replacing most universities with federated campuses, eliminating their managers, open classrooms, LLM TAs and other resources to respect the diversity of the student population, etc. would be better for the world, not worse. If you want to take the side of the incumbents in an economy already teeming with inequality be my guest.
The saving grace of the approaching compute as utility end state is the destruction of the class of interests you purportedly represent. With mass layoffs will come mass demand for UBI, and the destruction of universities that are in truth the ones doing the rug pulling you mentioned elsewhere. People complain about public incentives for AI? What about schools? Has the quality of schooling increased with price? And I'm supposed to be taking your side?
The ultimate rug pull is one we should all look forward to - the polity saying, thank you for your hard work Mr. Altman, Mr. Musk, Mr Whoever but this technology belongs to everyone now. It would only be right given we all know they are trained on a corpus that really should be considered a public good.
If LLMs become the "everything app" they will have to either be nationalized by the state (UBI), or the billionaires are going to kill us all with their robot soldiers. The battle is not over technology, it's about ownership. The owners you represent are less desirable because their continued existence is further behind the most desirable end state, and they themselves are only desirable if you assume only the worst possible futures will come to pass. I believe if UBI is progressively scaled enough over the years enough billionaires will be satisfied to refrain from mass homicide, as has happened before when wealth generating technologies essentially became state utilities.
Good luck in your efforts. It's always possible my model of the world is incorrect. A meaningfully informative rejoinder in that respect would be welcome.
I am not suggesting we don't adopt tech, it's how the tech is allowed to be sold and propagated through society. Using your analogy it's like if we allowed only a single company to manufacture cpus, hard drives and build the software such that there was no possibility of a competitive market. That's a policy choice that can and should be evaluated for the harms monopoly systems create. We as a civilization don't have to allow openai and anthropic to move into the app layer we can say as a condition of the ability to operate a commercial enterprise they are prohibited. We do that with antitrust law, utility regulations, telecom and countless other sectors.
I am not advocating protecting the management class, quite the opposite, I am saying their demonstrated greed in vertical and horizontal integration should be stopped because they have shown they cant be responsible with management authority. You use the word balkenized to imply that fragmented AI is a bad but fragmentation is the basis of free market capitalism and in particular the start up ecosystem.
Your advocating that society just capitulate to Sam Altman and grant a permanent monopoly and the status as permanent underclass because he did "hard work" but that is flawed for two reasons: 1. Hard work doesn't justify destroying the very system that allowed the conditions for that hard work and innovation to create wealth by letting winners convert a free market to a monopoly. That tension is the internal contradiction of free market capitalism, winners will become monopoly such that no more winners are possible. We solved this in the progressive era with utilities, oil companies etc, by either breaking up the companies that had too much market dominance, or making them divest vertical integration units (like standard oil having to sell its rail lines because owning transportation gave it an insurmountable advantage). 2. There is no AI without training data AI which was taken without consent or compensation from public and private sources created by civilization over generations. So the polity has a clear preceding claim to that work and if any industry lacks a right to the ownership of what they've created it's AI. Sam Altmans hardwork is trivial compared to the totality of civilization writing over generations and creating the internet and maintaining libraries and a mailing system so they could collect that data economically.
For those reasons it's totally justified and we'll within the legal and commercial precedents we've established to limit how a single company can develop and operate AI. We should not allow vertical integration across the stack. Openai cant manufacture chips, train models, and build an everything app from customer training data. I don't know ow what owners your think I am defending or representing but I would suggest you are arguing with a phantom opponent that isn't me or are misunderstanding my position. I am not suggesting nationalizing AI or giving the polity ownership, I am suggesting standard corrective regulation to preserve a free market with clear preceding examples from the last technology transformation. We suffered great harm when the electric grid and phone lines and oil and shipping consolidation was unimpeded and unlocked innovation and broad wealth creation when we broke them up.
I appreciate the response but I think you need to read what I wrote again.
I don't think I'm arguing with a "phantom" at all, rather you seem to suggest I want OpenAI to own the economy when I rather precisely articulated the most desirable "rug pull" (using a phrase you yourself invoked elsewhere in this discussion) would be depriving them of that very opportunity. If anyone has misread this, I would suggest it's you. We simply seem to disagree on the timing and nature of the regulatory intervention.
Are you against UBI? We've gone back and forth with each other for a bit now and I still don't have an answer to that question.
I have read your posts closely. If I misunderstand them I am open to being corrected. So here is what I take from them. I don’t think you are advocating for OpenAI to be a monopoly. I think you assume that it’s inevitable that someone will. And the only solution is to nationalize them and then use the proceeds to fund UBI.
Is that accurate and complete? If not, what am I missing?
Btw you didn’t ask me for my opinion on UBI, you declared your opinion as a statement or at most a rhetorical question not something that warranted a response. That might not be what you intended but reread your sentence and ask if someone would interpret it as a question rather than a statement.
I am not opposed to UBI or nationalizing AI, and I am not defending a managerial or ownership class. I am arguing that there is an intermediate step available to prevent the emergence of a monopoly that is fully vertically integrated with custom chips —> training —> inference —> application—> deployment.
Society doesnt need to allow that. We can limit OpenAI and anthropics ability to build the app layer. We could limit their share of compute to a percentage of available compute and make the remaining share available to other competitors. We could require that they distribute their training data to others after some period of time. We could require they release their weights.
There are a broad set of tools available to ensure that don’t capture the whole intelligence stack and then expand horizontally to, as you said, make it so “my business doesn’t exist in 10 years”. The fact that you don’t know my business but could credibly predict that they could put most businesses out of existence is a massive problem that UBI won’t fix.
I don’t want to live in a society where humanity is living on the patronage of monopoly AI. And AI won’t be viable if no one has money because they greedily expand everywhere and put people out of work and businesses out of existence.
I think there are intermediate solutions before UBI and full nationalization are needed and we should implement them. Specifically, making sure vertical integration isn’t allowed, making sure compute access is broadly distributed rather than let openAI and anthropic lock it all up, making sure training data sets must be released within a short time frame like we do with patents, making sure the labs can’t copy customers.
You mostly have it right. I appreciate the response.
Now it is my turn to make sure I understand what you are saying. On that count, I think there is a remaining point where we are not understanding each other. And really the rest of the discussion hinges on this.
You say:
> I am not defending a managerial or ownership class.
But also say:
> I have talked to over 10 people at universities, SMBs, and large enterprises that are actively allocating millions of dollars to bring AI in house.
Who are these people with such authority if not the managerial or ownership class, or at the very least, their representatives?
I think what you are saying is quite plainly a defense of such interests. We sort of need to agree on that or the rest of the discussion cannot follow.
Once we align on that, I will have more to say which I believe will close the loop I have been attempting to weave.
I have been enjoying this back and forth though. Some of your points do make sense. Perhaps you are even correct. But let's be accurate about the nature of the two options we each propose, only then can we attempt to grasp some sense of the tradeoffs.
Where do you live in a hut somewhere? How are you accessing the internet?
Civilization is all about giving up control. Are roads also a mechanism to control people?
UBI is desirable and will likely only come about by the sorts of collective action that defined the labor movement. Are weekends also a "mechanism to control people"? I think you need to consult a history book, it would give you a better grasp of who is "spreading nonsense" here.
Don't worry, we had DreamWeaver, Wordpress and other tools for a long time now but it still did not replace web developers. Build quality tailor made websites that connect to humans and you will be fine.
DreamWeaver and WordPress both had limits on the complexity of site they could build. If you wanted more than what they can do, you look to web developers.
This was always the play…
Wait until the models are trained to work on “capsules of your data” stored by OpenAI, and deploy your apps/services onto OpenAI cloud where you can _easily_ share the development with your team members’ AI agents. Ofcourse, they will all work only if you use and pay for their models via their subscriptions. A new global {K-V, Relational, Vector} database that your coding model is already trained to work with! will come very soon…
They know the API price wars are a race to the bottom with little/no profit margins.
I think this is going to back fire. I have talked to 5 tech leaders at companies and universities this week about setting up their own infrastructure for inference and training. They are worried about multiple rug pulls: increasing cost of tokens, stealing ideas, training on data etc. they are all willing to uses less intelligence models if they avoid these issues. They’ve seen me setting up propelcompute.com for my start up and propelcode.app neither of which are intended to be commercial products and want to do the same.
If you're wondering why OpenAI's models seem to be struggling to stay ahead of Anthropic and are falling further behind, I think it's because of side quests like this.
They have dots, they had Sora, so many other things that take away attention from just having the best model. There are already a bazillion ways to do this. Sure, the integration's nice, but it's just not needed.
Weird to describe something they could really easily sell to non-coders as a "side quest".
I think programmers make the mistake of thinking generative AI is for us, when it is much more ominously about us.
Newsflash: a company that plans to be the most valuable business on earth and has a desperate need for revenue is not going to stay out of anyone’s lane if there is money to be found in it. If you can make an AI wrapper, so can they.
This actually looks like it has meaningful potential, and given how easily they could train it to connect to a whole bunch of off-the-shelf APIs running in data centres they control, looks like a little glimpse of the future, and of the kind of power the industry has gleefully handed to the crypto-for-your-eyeball-scans guy.
Cloud AI has been the face-eating leopard of the tech industry for the last three years. This is them just starting to chew. How long before they tweak it to generate intranet apps, mini stores, specific AI wrappers? Why wouldn’t a firm that has so much influence over data centre rollout, RAM and processor prices simply directly start using their code generation tools and data centres to build and run the work product?
Maybe some of the big firms who have been helping train ChatGPT to eat their lunch will now wake up to the need for open weights.
I think at this point assessments like this are more qualitative than quantitative. I use Codex far more than I use Claude for coding and my results are generally higher quality. I would not, however, go to someone and tell them that Codex is better than Claude generally. In large part it depends on how you use the model and the harness.
And even if you do know how to use the model and the harness, Claude is just generally still a much a better experience overall.
Returned to Claude as my main driver after many months of Codex and my frustration index is in much, much better place. I can't recommend Codex/GPT to anyone, not just normies. It does make for a really nice reviewer though.
All of these "side quests" (and especially this one) are attempts to build a walled-garden.
Especially this. OpenAI will host it. OpenAI will throw "Sign in with ChatGPT" there. "Apps" will be built that depend on end users ChatGPT "connections". The app will fundamentally not work without an end-user having a ChatGPT account/subscription and "connected apps".
Having "the best model" will absolutely not support a $1T+ valuation. Owning the activity of 25-50% of internet users (assuming a split between OAI/Anthropic/Google/maybe-Meta), is.
For example, here's a game where you are a mouse hunting for cheese in higher dimensional mazes: https://manifold-six-axis-escape.nathan338932.chatgpt.site/
Idea came to me while at a concert, was ready to play that same night.
reply