And unlike the traditional "this will replace humans right away", I think what this introduce is a lot of incentive to spread those token in places where there was never any incentive to hire a software engineer for previously. In turn, that will drive a lot of business activity in those area that will potentially fail given the current quality of the output.

This feels like a boom before bust scenario, and I'm not even sure if it will bust.

Seriously, what value are tokens providing other than justifying layoffs. Concretely. Today. Not in the speculating scenario that cardiologist could be replaced with models.

We see this new trend of agentic coding, again a promise software will be written that way going forward, despite the number of fiasco already experienced when trusting a model turned bad. The use case may provide value, but right now all it does is fullfil the push for token consumption all these AI leaders are advocating for.

All that said the dotcom boom is extremely analogous and that crash was quite bad.

On the other end we have professionals happy to pay a subscription for heavier use, to build something in the hope to sell it.

I figured I don't believe in value when my dad explained to me his mate fired his team once he realised he could just pay 20 bucks for his Gemini account and run his business. I asked, do you call this value add? He said it must be, since he can produce the same output with no staff.

There is a confusion between profiting from a circumstance and value creation.

You create value if, say, you cure a disease. That it takes you an army of staff or extract maximum profit from it is just a feasibility formula.

That you make the cure more affordable is value creation.

That you cure the same disease but increase your profit doesn't create any value, except to yourself, for a while

Maybe you don't, but it's fairly obvious that a lot of things are changing and things are moving.

Maybe your dad's mate didn't have to expand on his business, good for him. Other business are expanding because they now can.

Will the positive overweigh the negative? Not necessarily, but to go "it's tulip" is the kind of argument so devoid of nuance that we shouldn't be discussing so on HN.

The overwhelming demand for token would not coming from people wanting a unique illustration - it would be from professional usage. In fact, I'm not even sure who is subsidized. The $20 subscription surely isn't being used fully across all members of that subscription.

2000's tech bubble was caused among other things over-investment to infrastructure and technology that had no users yet.

Totally different setup.

Does not mean AI boom will not turn to bust, but weak analogues generally don't help with understanding complex systems.

Claude helped me implement a ridiculous amount of features in my programming language. It's helped me migrate the heap to an easily moveable index-based object space. It's helped me implement generators. It's helped me implement a new memory allocator. It's helped me fix a ridiculous amounts of bugs and make a huge number of small improvements everywhere. Its ability to provide me repository wide code review was a game changer for a solo developer like me. And it's doing so much more than that. I got more things done in the past few weeks than previous months even though I'm evaluating, learning, understanding and rewriting the AI output.

It's actually addictive to build things with Claude. The usage limits are starting to make me anxious, just like withdrawal syndrome. I applied for their open source max subscription program even though I'm too small for it because who knows, I might get in anyway and it costs nothing.

AI is quite literally a world changing technology. I hope the open models keep steadily progressing and that hardware remains available to all so we can run our own models on our own computers one day.

Just far cheaper (if you are in USA) and probably more useful in terms of job prospects.

Agentic coding absolutely blew up from demand, users are not being tricked into paying $200 a month, and they’re not complaining about hitting rate limits because it’s useless.

  users are not being tricked into paying $200 a month

This belief is so common amongst AI collapse people online. I'm guessing these people have only used free ChatGPT or worse, they use Windows and get Copilot shoved down their throats?

Meanwhile, I'm flying around with a $20 Codex subscription doing everything from writing code, analyzing stocks, coming up with ideas, etc.

IMO if someone tried this tech last time 6 months ago, or their only exposure is eg. via MS copilot, they do have a rational reason for skepticism. No technology of this complexity has improved this rapidly in my memory (well, ok, we had the CPU speed races from 90's to early 2000's).

From the 80486 to AMD Athlon64 X2 and much of that progress was enabled by better EDA being run on the more powerful CPUs being made with each improvement.

Now, we have better models helping to create even better models.

Let's say they have already plateau. But hardware continues to get better, right? So tokens should go down in price, not up. Since they're already 50%+ on inference today, better hardware would allow them to generate more tokens for less money.

I would pay $500 to start, build stuff with it, then keep going up the tiers as the stuff I'm building makes money.

It's adding tests for me and doing medium complexity refactors that I'd otherwise have to spend hours on

And based on reality (code) rather than my feelz of what I vaguely remember the code to have been doing in some long past.

there is an even larger force on HN that financially _needs_ the value of tokens to be inflated (so much so that bots have overwhelmed the site)

Like the OP said, it's incredible how polarizing this debate is. When I read comments like yours, I feel like a significant part of the global workforce in IT must be living on another planet? Or they never really used Claude Code, Codex, OpenCode, ... intensively before because of company policies?

I legitimately am at least 10x more productive than a year ago, and I can prove it in number of commits and finished monetizable features developed per day. Obviously my workflows still very much require an active, constantly context-switching human-in-the-loop, but to me there's absolutely no question both output volume & quality have skyrocketed.

There are millions of other wanna be engineers doing exactly the same, assuming demand will scale as much as the offer.

What returns are you getting on those?

Let me create 500 websites, deployed for free, I hand that over to you by end of day. Will you give me a cent per piece? If so, happy to do business with you.

I would happily pay $200 a month for this. Luckily I dont need to, it's free.

Literally every game and website that I would have had to pay someone else to make I can now make myself. There's no value in that?

A year ago the best free LLM couldn't even give me a basic gridmap and collision. Now it can give me a full RCT style prototype & editor in 20 iterations.

I can only imagine what improvements we will have NEXT year!

Ponder that for a minute.

There are over 2 million games, for Android alone.

That you weren't making games before the advent of LLMs makes it cool for you to build, and at no cost. But people have been able to make games without them and already grew the market to saturation.

If the outcome of LLMs is that we get more games, it won't imply that people will consume more games. Most games never get played anyway.

Op never said they're selling games. They said they're making their own games and websites for a fraction of the cost (even $0). That's amazing value. And it's just getting better.

I didn't mean to patronize, sometimes self evidence isn't trivial to notice.

That claim is totally worthless without you providing concrete information how you measured that.

Value add so far lacks evidence.

Layoffs. It justifies them to the public. I'm not certain it grants them as it contradicts a principle of enterprise: scale, as much as you possibly can.

If tokens provided value today, we would be hiring more engineers to review their output and put things together.

That number is at least tenfold of what it was before, simply because I can run a lot of gruntwork in parallel now without wasting brainpower and focus on that stuff.

Coding, writing, summarizing, translating, data analysis, customer support, test generation.

Maybe we are closed to singularity, or maybe we'll just plateau somehow. But either case there are so much work to support the breakneck change that are not done because the change take priority every single time, there should be a lot of things to work on.

The question is how big the fail is if you measure it in 3 month increments going back to late 2022.

As long as there are more amount of success, then it should be net positive.

> For the data center build outs, demand for tokens is still exceeding supply.

Can you provide any numbers for this?

Now we don't know the true size of any of the proprietary models, but my educated guess is that Sonnet is in about the same parameter range, just with better training and much better fine tuning and RLHF. Yet API pricing for Sonnet is $3/MTok input + $15/MTok output, exactly six times as expensive. Even Haiku is twice as expensive as Kimi K2.5.

I find it difficult to believe in a world where those API prices aren't profitable. For subscription pricing it's harder to tell. We hear about those that get insane value out of their subscription, but there has to be a large mass who never reaches their limits. With company-wide rollouts there might even be a lot of subscription users who consume virtually no tokens at all.

This is false. We may assume it's the most efficient way of generating revenue given their GPUs, but their overall profitability will just be a guess. They would still have incentives to run hardware at maximum, even when it's uncertain to eventually recoup costs.

> a world where those API prices aren't profitable

A lab with employees and models in training has other costs than the operating expenses of a GPU farm.

If they're losing money and have no VC backing, they'd just turn off the lights.

But that's moving the goalposts? The original claim was on inference itself, not the whole company.

> The cost to serve tokens is absolutely profitable today and that’s been true for at least a year.

Everything is profitable if you ignore the costs.

Are you sure? Surely there is a lot of interesting data in those LLM interactions.

It's fair to say that if all these operators are competing for tokens, that the OpenRouter token operator (not sure the exact phrase but the people running the models) are accounting for some level of margin.

However, how many of these are running their own data centers and GPUs?

If they are running their own infrastructure, then it's not a simple equation of if each specific token set is profitable, since it needs to account for the cost of running the data center. It could be that they believe that it is profitable in the long term by utilizing the long tail of asset depreciation, but that isn't guaranteed.

IF they aren't running their own infrastructure, then it's much easier to claim that it's profitable and has a margin (outside of running their servers to manage the rented infrastructure).

HOWEVER, a lot of data centers have some pretty crazy low prices for GPUs that may be vying for user base and revenue over profitability. In these cases, if data center growth starts slowing due to slower buildout then it's very likely GPU prices go up and inference stops becoming profitable for the open router owners.

So long term it's not clear how profitable even these open models are.

OpenAI and Anthropic definitely fall into the latter category too. Their infrastructure requirements are much higher than the open models, and they are being given huge discounts so Microsoft/Amazon/Google can all claim revenue (since they have profitability coming from other parts). It's not clear if OpenAI and Anthropic models would be profitable at inference if they were paying rates that cloud hosts would make a profit from.

There's just way too many dimensions to this scenario to flat out state that open router proves inference is profitable at scale.

That gives you a very good estimate of "how much can you serve the tokens of a model of the size N for while making a profit".

Now, keep in mind: Kimi K2.5 is 1T MoE. Today's frontier LLMs are in the 1T to 5T range, also MoE. Make an estimate. Compare that estimate with the actual frontier lab prices.

In the current volatile environment, the API prices are more of a baseline where we can assume it can't be much cheaper to operate these models.

For supply look at outages and growth rates at companies like openrouter. The demand is growing every week.

This is why switching to local open weight models saves a lot of money. (Even though it’s not apples to apples.)

This is the opposite of an AI bubble burst.

300k tokens for that hour.

OpenAI charges $6.

Those are pessimistic assumptions.

[1] https://lambda.ai/instances

“Technically correct. The best kind of correct”. So inference may technically be _capable_ of being profitable, but I have question’s about them being profitable in _practice_.

If not, you aren't breaking even.

(1) Rent your GPUs.

(2) Pay list price, no volume breaks.

(3) Get only 85 tokens/sec. Realistically, frontier models would attain 200+ tokens/second amortized.

Inference is extremely profitable at scale.

You're generating about 36 million tokens/hour. Cost of Mixtral 8x7b on Open router is $0.54/M input tokens. $0.54/M output tokens.

You're looking at potentially $38.88/hour return on that H100 GPU. This is probably the best case scenario.

In reality, inference providers will use multiple GPUs together to run bigger, smarter models for a higher price.

It’s insane

In the end value per user is what matters in relation to being a healthy going concern and valuation in relation to Meta for example. Value per token is what should matter too - after all that’s what people are paying for.

Currently on a given day I'm chewing through approximately the equivalent of my lunch money, but where there's opportunity to extract wealth, someone will find a way to do it.

The wealth of great open models provide an excellent base for fine-tuning, distillation, and RL. I see a lot of untapped potential in the field of bespoke, purpose-built models that can be served far more cheaply than the frontier competition. I would not be surprised if we see frontier-adjacent experiences running comfortably on a Mac Mini by year end.

With frontier models seemingly hitting diminishing returns in quality, I struggle to see a world in which gigantic, expensive, general-purpose models don't become increasingly niche.

He contacted the company saying he's not going to do that, to which they replied something among the lines of "sorry, that's our process".

But there is no real higher limit. Imagine a LLM which could answer the question "what does my company need to do to beat the competition?". And then realize that the competition asks their LLM the same question. So now everybody is bidding the price up or using more tokens to get a better answer

In that world there’s no reason for a business enterprise to exist.

Nvidia has a de facto monopoly in the datacenter-tier GPU market. He says this sort of stuff because he knows he can keep jacking up prices, because the cost will be transferred onto consumers - mainly software engineers.

'Overinflated' relative to what? You make some good points but I don't accept this as a premise.

Median senior SWE salaries in SF: https://www.levels.fyi/t/software-engineer/levels/senior/loc...

Median income in metro areas: https://www.cnbc.com/2024/07/11/the-median-salary-for-the-25...

Engineering salaries are significantly higher than nearly every other industry on average and on median. Much of this is driven by VC funding rather than sound, profitable, bootstrapped businesses with sustainable profit margins.

Engineering salaries have also been driven upwards significantly the past ~10 years (since the post-2008 crash recovery), while wage growth in the US is mostly stagnant. I don’t have a source handy for that, but there are plentiful studies.

Outside of the US this may be less true, but I took GP’s “most of us on HN” to mean people who work in US tech companies which are primarily concentrated in high COI areas.

There was a surge in demand for SWEs and scarcity brought salaries up. Are them too high? Hell no. On average, my colleagues and me generated ~2M$ each in 2025 for our company, while we get payed a fraction of that (grants and bonuses included). If you look at net income per employee we are at around 700k each in 2025.

Additionally, employers try their hardest to drive costs down (eg. offshoring as much as possible, everyone doing layoffs at the same time, ...) and average/median salaries remained high. If the salaries were overinflated those numbers should have came down I believe. The fact that they didn't makes me think that it still is a scarcity problem not an overinflation one.

So by that logic, housing in coastal cities also aren't "overinflated"? After all, like SWEs, they're they're also scarce and in demand. They're also providing enormous value to the people buying/living in them, otherwise they'd be living in Oklahoma or whatever and paying a fraction of the cost.

Is the same on the job market? I don't think so. I never heard any SWE saying "let's scare people away from a CS career so we can bargain for higher salaries". The opposite is true though. Companies participate in career fairs, pre-uni events to make people gravitate towards a CS careers, ... so with a higher supply each employee loses a bit of bargaining power.

Small excursus, this very fact was taken to the extreme in 2022 when everyone did layoffs at the same time despite the numbers being still great. If you put 300k people on the street at around the same time you can hire some of them for way less money as they now lost all leverage (since there are other 299.999 people waiting in line for a job).

now compare the profit per employee at tech (software engineering) companies and those industries..

But there are thousands if not tens of thousands where the profit per employee is minimal or negative.

I can’t find a source for all tech (the data wouldn’t exist for private firms anyway) but I think it’s telling to look at this list, scroll down to about the middle and look around at salaries you or your colleagues are pulling. Software revenues are certainly high but the industry is afloat because of these high margin businesses creating returns so that low margin businesses can exist. Without the massive infusion in upfront capital, very uncommon in other industries, it’s simply not sustainable.

Typically a market that’s buoyed by its top performers but has significant amounts of capital tied up in under performers is called “a bubble”.

https://www.trueup.io/revenue-per-employee

How can you possibly say that? Everyone knows that's not the case, these companies are losing money every day selling tokens. Revenue is not the same thing as profit.

This is why they were freaking out about DeepSeek just taking the trained model weights and slapping an interface on it.

Of course they are profitable if you ignore their cost to bring a product to market.

Yeah it should be factored in, but it’s a different set of implications for long term sustainability. They don’t actually need to test and optimize a new menu every day or week. If they decide to just stick to the same one longer they can get way more return from each dollar spent on development. It’s just that right now the rate of improvement you get with training is really high and nobody can afford to fall behind their competition.

Yeah, sure you can ignore the cost of purchasing the building for the restaurant for most profitability calculations, but if every year or two you were tearing down your old building and building a new one, you better believe that has to be in your profitability calculations.

If you can simply not remodel your restaurant and keep making money, then yeah, it makes sense to call it profitable.

Can you explain why you know better than the analyst at Cursor cited in this article?

> well most of us here on HN have benefited from decades of overinflated engineering salaries being paid by often companies that were not profitable and not only unprofitable

This is a really concerning perspective: people were paid what they were worth. Software is or was one of the few remaining arenas wherein a person can find a middle or upper middle class lifestyle consistently.

I will also note: a startup raising an 8 MM series A and eventually fizzling out is not the same at the hundreds of billions invested in these AI companies without a path to profitability. It is utterly absurd to pretend these are the same thing: any company ingesting that much cash needs to justify its capacity to survive.

Software salary inflation and expansion has made this the case. Tech’s accessibility to the educated has accelerated gentrification massively, rising up prices on rent and food. While the statement is correct, tech’s contribution to income inequality is part of the issue. If you’ve lived in Austin or Chicago (especially Austin) prior to ~2010 you’ll have seen this first hand.

Rent prices push everything up in the local market. Housing rents impact business rents in an area, as well as what the business’ owners and employees need to maintain their own lifestyles. People who live there now can pay more and will, so prices go up. But the local shop owner isn’t getting rich, they’re still struggling as everything around them rises.

The salary jab was probably a little harsh.

Your ending is a bit of a fizzle too. There are many capex intense businesses that do just fine.

I want to add something additional to this: it is one of the few fields that can afford middle or upper middle class lifestyle and is accessible.

I have no doubt if I could redo my life with the necessary resources I’d be more than capable of putting myself through med school and gone with a secure career that paid more than I ever made in software.

But at this stage of life? I don’t have the time or money to spend a decade+ paying some institution tens of thousands of dollars to hopefully maybe have a real career.

Once software as a career dies, I suspect many will find themselves locked out the middle class for generations.

On the other hand, once software as a high paying career dies there will be nothing to prop up the status quo (high cost of housing, for example) so the middle class will return to being much more accessible to modestly paid jobs.

What, why? There are tons of low-margin capex-intensive business out there.

I think AI will end up like being like hosting. All the models will converge to being pretty-decent and the companies will have to compete on efficiency since they are selling a generic commodity.

You can already see Anthropic fears this scenario since they try so hard to make people use their first-party tools rather than plugging Claude in as a generic part of a third-party stack.

LLM hosting is the next VPS.

Even interpreting what-they-were-worth in the usual sense, I’m not so sure about this. We have seen wage collusion reported by the usual US West Coast-based companies. And some news on here[1] have reported that some engineer with a salary of $100K[2] might be producing $1M of value. And even factoring in the usual “but benefits and overhead” comes out to a solid factor of profit per programmer/engineer.

Despite that the sense I get (only from this site since that is my only reference) is that the so-called overpaid engineers are incredibly content to just have this happen to them. As long as they are paid well compared to other workers, it’s fine. No matter the profit factor. In fact, the discourse is very much focused on how “privileged” they were if the tide ever changes. Instead of realizing how much value they provided, collectively.

Outlets for capturing more of the value they create is entrepreneurship (Hello HN). Never any collective organizing. And entrepenurship is easily bought via aqcuisition.

Collective bargaining would have been relevant in case they ever get automated... by the very software they co-created.

One could imagine that this “privileged” collection of programmers could have served as a vanguard for the collective good of programming professionals as well as collective ownership of software goods, using their privilege to that end. The former never happened, and the latter is partly realized in people’s free time (see the OSS maintainer in Nebraska meme).[3]

[1] All from recollection since this is just news from the Frontier to me

[2] Of course the pay might be much higher now; this might have been a while ago

[3] when it isn’t simply exploited by corporations just using OSS without giving any back; a logical turn of events when no license or law forces them to contribute back

Well I’m sure they’ll be thrilled to know they can collect $100 a week more in unemployment benefits than their neighbor.

The parent comment doesn't discount that, only pointing out that "what they were worth" was inflated due to a speculative environment. Wherein lies your concern?

Salaries across industries in the US have remained flat since the 1970s. Calling the one sector that can provide access a middle class lifestyle inflated s to play into a narrative capital is eager to tell, even if OP didn't intend that.

What do you mean? The real (meaning adjusted for inflation) hourly wage in the US has increased by around 20% since 1970.

What has changed since the 1970s is that wages are no longer coupled to productivity. Perhaps that is what you are thinking of? But that should be an obvious truism for anyone in tech. We create the very things that cause that to be the case!

What happened in the 1970’s was the NeoLiberal shift and wasn’t caused by software.

If we — those with the power to build the productivity creators — took a stand and said "we refuse to create tech for the interests of the few" it would have never happened. But, instead, we welcomed it and are responsible for it.

So no. It wasn’t caused by tech beyond the uninteresting factors like modern society being complex and, of course, that tech developments influence things (pretty much all things).

The benefactor of those gains was also entirely decided by those who created the tech. We could have given use of that tech to everyone. In some cases we actually did (e.g. open source), but in most cases we gained (at least partial) ownership of the capital so it was in our best personal economic interest to keep it for ourselves and our close friends.

Would have been impossible without and being caused by are different things.

The sense of being “caused by” in a political context are the people who have the power to direct things. Which are not necessarily the people who implement something.

> The benefactor of those gains was also entirely decided by those who created the tech.

You assert that they were decided by. Based on what?

The vast majority of tech work was done in employment, either for some government or for private entities. The private entitites were controlled by Capital. The governments were controlled by democratic forces and Capital.

> We could have given use of that tech to everyone. In some cases we actually did (e.g. open source),

Again I reference the meme of Overworked Nebraska OSS Maintainer.

The impressive work done on OSS by tech workers directly have been done in their free time. The bulk of OSS work done by people as a living is probably through corporations like e.g. Intel working on the Linux Kernel.[1]

That impressive free time work has gotten the reputation as a treasure trove for the highly motivated and tech literate. In contrast to something that regular people can plug-and-play as an alternative to Big Tech dominance.

> , but in most cases we gained (at least partial) ownership of the capital so it was in our best personal economic interest to keep it for ourselves and our close friends.

Yes, well played. For those that got away with their financial-independence millions. For the rest, well, I guess they never managed to learn the moral lesson of Monopoly.

[1] Or am I wrong here? I could be off-base.

While you are right to recognize that there was some attempt to inject political context, it was not there originally, and is not the main discussion taking place. The fact that wages and productivity have become decoupled is not inherently political. It is but simple mathematics. Tech is the cause for the decoupling; it is why we have been able to become continually more productive and at an accelerating rate.

> The vast majority of tech work was done in employment

Yes, but generally even where employment is present tech workers also demand a share in ownership (e.g. stock). Tech doesn't invent itself. At least not yet. The workers have held the cards. Even those who haven't won the lottery are still in a pretty good economic position, relatively speaking.

I don’t care if anyone wants political context to be there or not. Political context is not some subjective choice that the participants in a discussion can choose to be the case or not, like some alternative history exercise.

This political context (i.e. reality) called NeoLiberalism is so well-researched and argued that I can just call it NeoLiberalism and even a forum full of techheads don’t bat an eye. Which is more than can be said than your incoherent nuh-uh where both:

- Technology just determines things by itself

- And the rank and file peons who implement technology could have forced something better on the world (than the pile of shit that we have)

“Inflated due to a speculative environment” is not an accurate way to frame labor prices that held for many years. At that point, the prices were simply high due to high demand relative to supply (compared to other types of labor).

That goes without saying. The investigation here is into demand. Which was said to be overinflated due to speculation. As noted, many of the companies hiring the developers did not have viable businesses.

In accounting, almost anything you want can be true, at least for some time.