AI will end up being a commodity. That's also why you already see the first rounds of protectionism under the guise of national security. Obviously they do have a point but that point works against them just as easily. Anybody that integrates unreviewed model output into their workflow is asking for trouble.
non frontier intelligence will be a commodity. Frontier intelligence will be a high margin luxury good. this is how most markets work
Yesterday’s frontier is today’s commodity, so I don't think frontier AI is out of reach. With computing and solar energy costs continuing to drop, the current AI supply shortages are just a temporary blip. We’ll probably land right back in standard boom-and-bust cycles.
Either frontier AI will continue to get better steadily, until it's better than people at basically everything, and very weird things start to happen, or, it plateaus and the open weights models will eventually catch up.
There's a third possibility, which to me is also the most plausible. Models will keep getting better and better within certain domains, while never catching up with humans when it comes to other type of tasks. Like that elusive quality-without-a-name some have called "taste". Or knowing what UX is most intuitive to humans. Really, anything where we have to tap into our common humanity and our human experience in order to be of help to other humans and generally to life on earth. Which is an all-important skill, since the flowering of life on the earth is (or should be) the whole point of anything we do, after all.
We've spent a couple of millennia documenting that elusive thing called 'taste' in word, sound, image, lately in moving images. I'm pretty sure that even if you didn't have taste you could make a good simulation based on all that input.
I don't know, that sounds like a question of taste. :)
It can remix things people have already documented but can it synthesize an entirely new thing?
to synthesize an entirely new thing you have to start form the big bang. Everything else is derivative.
So what you're saying is that since there are so many examples of what humans have preferred in the past for these models to be trained on, surely there will come a time when they will know even better than us what we want and what we like.
... as long as they still care about what we want and what we like.
Everyone knows the path followed by every single technology mankind has ever developed - including the incredible run that was Moore's law. Sure, some will think "This time is different" - but is it?
It doesn't need to run indefinitely. I don't think anyone sane thinks it's going to run indefinitely. It just needs to run long enough. Moore's law was what, 50 years? If AI keeps improving at the rate it currently is for 20 years, the world will be very different when that curve flattens out, in a way that makes most of our current concerns are entirely irrelevant.
Those cycles have shortened considerably. Just look at the lifespan of media, integrated circuits, software, products. There have been so many big model releases just in the last three months I've lost count. We're now closer to weeks between steps.
Maybe if China stops releasing capable models the pace will slow down a little, but they have lots of incentives not to.
It is, because just like you mentioned, there's no exponential physics to back up growth. LLMs have to stop scaling up compute, because there's only so much resources.
But even if we think back to the previous decade or two (sans AI), most of tech has been a a red queen's race - developing and improving stuff, without any real substantial results.
Microservices and orchestration are not better than old monoliths. Electron apps are not better than native. Most programs people actually use have decades-old codebases.
Video games of today don't look substantially better or play different than ones from 15 years ago.
But it is not a constant point of "this % of people needs the frontier intelligence".
As thje models get better, the % of the pie gets smaller and smaller. So higher and higher cost of getting frontier model out gets divided onto smaller and smaller customer base as other models catch up
Yes, this is the crunchy bit and we're already well underway on that curve. I haven't touched a frontier model in months because I can do with smaller ones without wondering/worrying when I'll get banned next because what I'm doing is dual use. Qwen was the the first time I thought oh, ok, so this is 90% of what I need. Then DS4 made that 95% and with GLM 5.3 it's somewhere between 99% and 100%. Now, I'm not trying to solve any Nobel prize winning stuff here but if that's the kind of customer that will remain I don't think the frontier model providers are going to like the next six months or so.
That's the crux of the issue. You think that GLM 5.3 is all you need.
However, the expectations for AI are not to make your software engineering work 100% faster.
It's, for example, to build software more or less end-to-end with minimal input and feedback from you.
Then next, robotics and boosting manufacturing. Not by 10% but let's say 200%.
Such are the ambitions. Can GLM 5.3 do it?
The crucial question is whether eg. Fable 6 can, and with what lead over the competition.
Open-weight models still have not beat February's Mythos.
I am actually doing robotics stuff and yes, GLM 5.3 can do it.
As for the manual feedback: I want that manual feedback, but if I let it rip it will do just fine all by itself, but then I have no clue what is going on under the hood.
At some point you have to let go of trying for full understanding and control, delegate to the AI, and unlock the real productivity boost.
I have no clue about the state of robotics.
For software, I know that Opus 5.5 can handle more than GPT 5.6 Sol. I am willing to pay for the difference rather than use the cheaper DeepSeek Flash.
> At some point you have to let go of trying for full understanding and control, delegate to the AI, and unlock the real productivity boost.
The prerequisite here is that you can trust it not to do something that gives criminals access to your bank account or allows users to consume an unlimited amount of services it costs you money to provide for free or bankrupts you through legal liability etc.
It's basically this:
"A computer can never be held accountable, therefore a computer must never make a management decision."
Notice what this is actually saying. It's not saying that the computer can't generate a given output. It's saying that if the computer makes a mistake and you don't bother to check, you're the one who has to eat it.
> "A computer can never be held accountable, therefore a computer must never make a management decision."
That's a great way of putting it. So far I always used 'you can outsource work but you can't outsource responsibility'.
> "A computer can never be held accountable, therefore a computer must never make a management decision."
Legendary quote from an internal 1979 IBM training document [1]
https://simonwillison.net/2025/Feb/3/a-computer-can-never-be...
I'm getting plenty of mileage without that and anything with a motor in it requires that you do your utmost best to ensure it works within the design envelope.
We aren't going to get useful humanoid robots while understanding the implementation fully.
If you can't leverage AI to build a safe and useful robot without fully understanding the implementation, the model is not good enough.
There are so many assumptions in these two lines that I don't know where to begin. I wish you best of luck.
The humanoid robot was an example.
My argument is that if safety and liability make the use of AI without fully understanding the implementation impossible, the model is arguably not good enough.
A long time ago we decided as a society that in order for an engineer to sign off on physical designs, they have to take legal liability for them.
No such process exists for a model to take a design seriously that can kill people.
For serious engineering work not only physical designs, software also.
When we manually tested nuclear power plant controlling software, each test case was printed out on paper and tested and then signed by the tester. A lot of paper was produced and archived for each software release. (Archived for as long as the software was running in any nuclear power plant).
Everyone understood, that in a case of significant accident at the nuclear power plant, accident analysis would be done and if the root cause was an software error, this could have possible legal consequences for the developers and testers. (Of course nuclear power plants have many independent and redundant hardware controlling mechanisms, but incorrectly working controlling software could damage the power plant in a very costly way).
With a larger project, I do not think there is a requirement that the engineer signing off understands every implementation detail.
I imagine that the requirement is that after testing, you are confident that the product is safe.
Respectfully: you are out of your depth. Maybe pick another subject?
It was you who shifted the conversation away from software to make an argument that "anything with a motor" requires full understanding of the software implementation, and that for this reason GLM 5.3 is good enough.
I think you're entirely wrong, and we can leave it at that, seeing that you're clearly unwilling to engage constructively.
The practice of deliberately engineering software without full understanding of the implementation, especially when it comes to robotics, should be illegal. Technology serves society alone and it should progress at a pace that is sustainable for adaption. Your excitement for a “singularity” does not outweigh someone else’s health and safety.
Nah man, no you don’t. There are very few absolutes in the world and that certainly isn’t one of them.
And over a increasingly small slice of time, as, due to the commodification of these things, the frontier businesses will see their customer base walk to the next one that does any better on their task.
This is probably ultimately unsustainable.
Every time they throw someone out with a message about how this request is definitely against their safety guidelines they are clinching the deal on someone running something local or switching to a provider that just lets them get on with their day. Their own fearmongering is going to be part of their undoing.
A high margin luxury good with huge r&d costs is a death spiral.
A high per-user r&d means high costs per user. So then users leave to save a buck, qnd the price goes up again for the remaining users.
Maybe it works out OK if you want to signal wealth - the high price is the selling point. But luxury pure utility is a hard sell to consumers, and for luxury B2B it's about high touch sales (e.g. Oracle).
Agreed, but I I wonder what fraction of tasks really require frontier performance, either in terms of sheer difficulty or where speed is a matter of necessity. For many jobs, it's increasingly a question of whether you'd rather have one filet mignon or 20 good cheeseburgers (all from the same cow).
I’m not sure if this counts as a nitpick, but the definition of frontier automatically includes the cheapest models
Who can and is willing pay top dollar for frontier human intelligence, and fully utilize it to actually make money?
The only answer I could come up with is hedge funds. Everyone else fails some part of my criteria.
The models will be a commodity, the resources to run the frontier will be the luxury good. This is already the case nearly.
The market for "frontier intelligence" will be too small to serve outside of certain government/military applications that exist in spite of market forces, in the same way that the market for small, portable nuclear reactors is exclusively limited to submarines and aircraft carriers. When I call a plumber, I don't need a frontier plumber, I just need someone who's baseline competent.
> market for small, portable nuclear reactors is exclusively limited to submarines and aircraft carriers
SMRs are being commercialized. Likewise nuclear cargo ships have been seriously studies. Nuclear reactor powered (not just RTG) satellites have been built.
I know this is beside your point, but I think if we figured out how to commercialize this, demand would come.
> I just need someone who's baseline competent
But why wouldn't you want (and pay) for the very best plumber? A frontier plumber will optimize the time their customers poop, pee, bathe, or do dishes 10x to 200x, and with their freed-up time,they'll outcompete those who don't pay for frontier plumbing!
You forgot the /s
Sir, I take the impending plumbing singularity very seriously, and wouldn't want anyone to be left behind. When plumbing hits recursive self-improvement, we may see toilets that shit themselves more efficiently than any man can, and finally attain ASI (Auto-Shitting Infrastructure).
I can see you do. I beseech you though to think about the consequences for other industries. For instance, the ratio of toilet paper:organic product will need to be maintained somehow and I suspect that if we follow your pathway to its logical conclusion that there will be a cellulose crisis unlike anything seen in the world since the USSR went bust.
The next move in my investment strategy is obvious though, invest to the hilt in PVC tubing and I thank you for this invaluable insight. Fortunately this forum has no visitors so we can rest secure in the fact that the pigs won't realize it is too late to avoid being slaughtered. Power to the bulls, after all it is their product more than anything that is responsible for the plumbing revolution.
there is definitely an asymptote
Non-frontier intelligence has become a commodity already.
And the success of things like ESP32 shows that good price + good enough is a powerful combination.
> AI will end up being a commodity
what does this even mean, the commoditization... like a thing you can buy from a gas station? AI commoditization is the rock bottoming of pricing.
Usually we end up with cartels in such cases, see OPEC, mafias etc. But we have China throwing sticks into the wheels of the US providers that are attempting to make the cartel and "pacing the frontier"
> Usually we end up with cartels in such cases, see OPEC, mafias etc.
Whether a cartel forms is independent of whether a product is a commodity. To the extent that commodity markets are more likely to form a cartel, it's only because it's harder for them to form a monopoly.
The general problem is that if your legal system allows competitors to buy each other then they have the incentive to do that instead of competing. This is why there are supposed to be antitrust laws, and why we get problems whenever they're not actually enforced. But in a typical commodity market, even buying the competition doesn't quite work because someone new can always enter the market. Congratulations, you've cornered the market on landscaping services... until some unemployed fellow who already owns a lawnmower and a truck starts advertising their services.
OPEC is kind of a special case because the product is entirely fungible but it's also finite and consumed by usage. Landscaping services aren't meaningfully finite, if demand increased then supply could match it under any plausible scenario. Copper is finite but it isn't destroyed by usage, so if someone had a lock on copper mining then people could still get it by recycling existing products.
Whereas oil is finite but then the end customer burns it. If Saudi Arabia cuts production, Germany can't make up for it by extracting more domestically because they haven't got enough in the ground, but they also can't capture the exhaust from their cars or power plants and turn it back into petroleum because the product is really energy and it takes more energy to do that than you get from burning it to begin with. So if the largest exporters get together, they control enough of the market to control prices and there is no real way for anyone else to enter the market with enough production capacity to change that.
Until you realize that the product is really energy and you can get out from under their thumb by switching to electric cars and generating electricity without burning fossil fuels.
A commodity is an undifferentiated good. I expect gasoline to be practically the same at each provider. AI will be similar.
The funny thing is the most lasting contribution OpenAI makes to this field could very well be the protocol for the endpoint.
Not too long in the future, I fully expect OpenAI to rescucitate the API copyright aspects of Google v Oracle to pad its revenue in OpenAI vs Everyone profitable
I wish them good luck. I don't doubt that they will try.
China is doing all the heavy lifting for the free market in AI right now. I did not have that in my 2020s playbook.
> what does this even mean, the commoditization...
Like a utility. Not 'ACME intelligence' but just 'Intelligence'. With multiple providers selling you intelligence that is 'good enough' by the token.
>Usually we end up with cartels in such cases, see OPEC, mafias etc.
No we don’t. That’s quite rare.
Look at the collapse of the solar market around 2010 as the price collapsed.
Look at the collapse of the steel industry in the US as outsourcing to Asia took over in the 70s-80s.
Look at the entire computing boom from 1990-2020. Memory and storage followed such a steady line down and to the right that people are struggling to operate compute businesses in a normal market.
E-bikes, scooters, electric cars, lithium ion batteries in general. Most food.
The list of these commoditized things that started out as areas of innovation and became just the marginal cost of production is enormous.