orangecat
orangecat on Hacker News
They're net negative by an order of magnitude on most of their current customers.From the S-1, in the first 3 quarters of 2026 their revenue was $33.2B and compute costs were $18.8B. Inference is highly profitable, they "just" need more of it.
it is the content provider who is paying the taxSee "tax incidence". Sales taxes are paid by businesses, and yet.
No LLMs for finding bugs.As of late 2026, this is gross negligence and any project with such a policy should be assumed to have a variety of easily exploitable vulnerabilities.
Yes. Also exercise reduces the risk, and COVID has plausibly caused a long-term reduction in exercise levels.
Good piece, and ditto. Really the worst part for me was just not being able to eat for 24 hours before; the actual procedure was painless.
So the food got better from 2000 to 2022 and only since then has gotten worse?
"They've specifically avoided trying" is also a possibility. Because what do they do if they actually find a critical flaw? Publishing leads to mass chaos. If they don't publish but it leaks, also mass chaos. (Probably just leaking the existence is enough without the actual proof). If they're "responsible" and report it to NSA, there's a good chance that model development gets forcibly halted and some chance that the entire company gets seized on national security grounds.
But this is the age when large government is generally seen as undesirable or a threat to the individualDonald Trump would agree that the government should have a large amount of power.
The correct takeaway from this is that the AMT should be reformed, not that seizing large portions of founders' companies is fine.
that would still only account for a 6x increase in revenue-- not the 10x increase that they need to fulfill their contractsAssuming you're talking about the $600 billion or so over the next decade, they don't need a 6x increase to cover that, they're already pretty close based on the $65B run rate as of last quarter.
their revenue is super concentrated in just a few customers
20% for the top two customers is not actually that concentrated. If both of them go to zero, at their current growth rate (45% from Q2 to Q3) it would take about 6 weeks to replace that revenue.
If the rich guy loses $1000 and you lose $10, is that good?
Did it catastrophically flop so the cinemas are ditching screenings?Yes. This story is a perfect example of being deceptive while not saying anything that's false. They conveniently leave out the abysmal revenue numbers, hoping readers will assume it was due to censorship or threats.
I think we'll see a lot less people willing to share if the end result is just to feed the machine no matter how they arrived at it.I don't see why I should care about this any more than I cared about "somebody I don't like could read my code and learn from it" in the past. The frontier labs can generate or buy all the training data they need; not releasing your app isn't going to hurt them in the slightest.
Give it another 6-12 months until the big AI shops shut those capabilities down and running local ones becomes cost prohibitiveI don't expect that to happen, but I do expect Apple et al to spend trillions of tokens hardening their platforms against users and red-teaming to find these sorts of holes before release.
People enjoy different things. Its not strange.Yes! Which means that the argument frequently seen here of "AI is terrible because everyone hates using it" is not valid.
Interesting, reminds me of the Stroop test where people have to say what color a word is written in, except the word is "red" and the color is blue.
I wonder if a sufficiently smart LLM would recognize the difficulty and build a preprocessor that would let it mostly work in a normal syntax.
Ok, can you list which diseases and conditions we should not be trying to treat because evil rich people would benefit?
Correct. To be fair, housing costs are out of control in many areas. But this is almost entirely due to governments preventing housing from being built, so it's baffling to see so many people blame this on "capitalism".
I understand that sentiment, but I don't think it really counts if the only reason you were able to do something first is that others were forbidden from using an AI that would have done it faster.
We do for seniors, who are much more expensive than average due to medical costs. In a hypothetical future where AI is much more productive than humans we could easily extend that to everyone.
This is not a reasonable take in 2026. The proofs are valid. You can even check the Jacobian counterexample yourself with pencil and paper.
Yes, the classic marketing ploy of making false reports to the FBI.
Seems like it, and that's probably not a bad thing. Most companies have no need for git's distributed functionality, and even most open source projects just treat Github (or whatever) as the single source of truth.
I feel it’s the ultimate hubris to think we can cheat death.And if we were meant to fly we'd have wings, and diseases are the gods punishing us for our sins.
Moreover, why would you want immortality?
I don't know about immortality, but I'd definitely like more than 80 years, and I'd like to be in good health for as long as I am alive.
Many books of fiction and poetry discuss the curse of immortality.
Yes, and those are almost entirely cope.
Right, disallowing AI-generated drive-by PRs is completely reasonable; almost all of them should have just been bug reports or feature requests. If your AI was able to one-shot the code, the maintainers can do that themselves and it's going to be much easier for them to work with their own AIs rather than using you as a meat proxy.
Uber is a vast improvement over the previous taxi oligopolies.