Fucking crash already.
If you invent something people want, they’ll buy it. If you invent something nobody wanted or asked for then you’re going to have some trouble.
It’s also a product that is more expensive to produce the more people actually use it.
I don’t know if it’s still going, but they were giving out free first month sub recently. You can grab it, cancel immediately, and cost them a few hundred dollars in tokens over the month using Codex.
Oh hell yeah.
We lose money on every sale we make… but we make up for it in volume.
“The money we lose increases volumetrically.”
I mean, that was the joke for Amazon for years, but…look where they are now.
But I just don’t understand the massive hype around LLMs. The cost per 1M tokens has dropped, significantly, since 2023 or so. And I don’t think there is any reason to expect it not to continue. If you host it yourself, it’s basically free other than hardware and electricity.
Sure, with agents like claude code, you may use a bit more tokens or use it more often, etc, and the best models are more expensive. But not every task needs the best model or even a paid model at all. And people seem to be finding ways to fan out things to prefer local models for simple stuff, escalate to something paid, etc.
I just don’t get the business model at all, but maybe I’m just too stupid to understand. It’s also possible we have an Enron-style situation, but spanning multiple companies and/or industries…
Amazon was building a platform, that now they use as a rentier. Bad for society, good for Amazon.
But LLM models don’t have that. You can swap one out for another with little effort - hell, I do it daily. And with that, it means that you’re a commodity . . . which also means you can’t ever get those initial losses back if your competitor (like Chinese models) doesn’t have them.
Have you seen openrouter.ai? They are more or less based around that idea, I think. Routing to whatever is best fit - most likely on price, I’d think (?) and/or currently up.
I have not set it up myself, but I think there is a way to have opencode or the like prefer local models for certain things, then escalate to something remote (presumably paid) if needed.
Also, opencode.ai gives you so many free tokens per day on some of their models.
The cost per 1M tokens has dropped, significantly, since 2023 or so.
The problem is that it’s just not commercially viable, because you can’t just say “We make money on inference”, because that’s a stupid thing to say. That’s like saying the cash register at the supermarket is really profitable, because you sell a product for a higher price than it costs you to buy it. It may be true, but that’s not how any of this works.
You can’t just stop developing new products, because if you do, your existing product becomes worthless. If Ford just started selling their current models, they would make a lot of profit right now… and then none in a few years as their competitors sell newer, fancier cars. If OpenAI stops training new models, they will make a lot of money right now, and then nothing in 3 to 6 months from now.
Token pricing could double, and be free for OpenAI, and it still wouldn’t save them, because you can’t judge a business on how much money they earn at the cash register.
I just don’t get the business model at all, but maybe I’m just too stupid to understand
There is no businessmodel. The trick to keep burning cash, keep the stock high, keep the compensation for the people at the top flowing. The company can burn all the capital it wants, because the people at the top will keep getting paid.
One layer below that, there’s a whole bunch of business idiots who think LLMs will let them save on paying their employees, so they’re very invested in it.
One layer below that are a bunch of senior managemers who are just as stupid, but whose performance goals were just tied LLM use, so they’ve got a great reason to push it on everyone.
If OpenAI stops training new models, they will make a lot of money right now, and then nothing in 3 to 6 months from now.
A problem the 2 big labs have, with OpenAI the bigger problem, is forced compute leases. The training is basically free because they have way too much GPUs already rented.
I mean, that was the joke for Amazon for years, but…look where they are now.
indeed, i refused to buy stock because it was costing $1 to make 80c in sales, what sort of business is that? and yet here we are, to be fair their cloud business is the big profit centre and that came later
Just pop already FFS
OpenAI is starting to look like the world’s most expensive experiment in proving that revenue and profitability are two very different things. The technology is impressive, but if every new customer brings another mountain of computing costs, scaling up might just mean losing money faster.
According to their own plans they expect to lose several hundred billions between 2026-2030, but revenue will catch up to the losses somewhere along the way?
Yes, it’s under step 2: “A miracle happens here”
They will never be profitable, it’s completely obvious at this point.
good
$10b revenue for quarter ($40b annualized), with claim of $50b annualized for september (not october claimed in OP)
It is also having massive fundraising problems, getting $11b-$15b out of targetted $30b, covering about 1 quarter of losses ($10.5b this quarter up despite about 50% revenue growth over last quarter). It’s forced compute spending increasing a lot next quarter, and then it pays Oracle $15b/quarter extra for compute in 2027. It’s optimistic gross margins on inference at the end of the year were 50% even as they cut prices.
Industry growth is slowing overall. OpenAI would need to average 20.3% revenue growth/quarter to break even at end of 2029. That is higher than their projected 2029 revenue, which with $250b/year in forced compute purchases then, would still leave them burning $70b/year.
There’s no obvious escape for OpenAI.
Thinking the whole industry is preticated on the assumption of a government bailout in exchange for unprecedented surveillance.
We gotta see if any of the CEOs running these companies are connected to old money somehow (either in the US or overseas). Why can’t it ever be nice people that are running the show?
Just as DoD harrassed ANT over full skynet permissions, you don’t need old money/old politician control vectors, the bond villains just need to understand that life is easier (well paid) for them if they cooperate with skynet.
What do you consider old money when it comes to the internet age? For the last good while it seems all the money surrounding PayPal/Oracle/Amazon seem to be the medium age money now. Old might be considered Jobs/Gates.
That said Powell (Steve Jobs Wife) was appointed to the board of Skydance last month.
Bugatti oil changes ain’t cheap!
I hear it sits around most of the time. Very selfish of the owner.
deleted by creator
OpenAI did not report earning $20B short of what it had reported a month earlier. It reported earnings $20B short of what one analyst had estimated, and the market had taken as broadly accepted reference point. Very different.
When you’re building an ever higher building by excavating the foundation to keep stacking things on top … eventually, you’re going to run out of material.
Like Jenga!
🫧🫧🫧
