This article is just kind of a mess. Yeah, if you ignore all the problems, the problems go away.
Anthropic is probably the worst company to look at because they lease all their datacenter space. They don't carry the debt, they don't play the games with depreciation, and they're probably most likely to survive because of that. Detailed analysis of Anthropic's S-1 in isolation tells you almost nothing about the problematic areas of the AI sector.
And there are a few really comical statements that render this whole thing unserious:
> The reason why AI companies are spending so much money is to rollout data centres to run the AI in the first place. These expenses are one-off purchases, not regular operating expenses. Once all the data centres are built, AI spending is going to plummet.
Okay, shares of NVDA just crashed to a round-off error close to zero.
There are other reasons why that statement is not accurate, but clearly the author hasn't even bothered to think through their argument at all.
> In a pinch, AI companies can also cut costs by no longer allowing people to use their products for free and temporarily suspending expensive AI research. This can be implemented instantly, whereas other kinds of expenses like leases and spending commitments are a lot harder to cut.
- AI companies are limited in their ability to jack up prices by open models being good enough and inference being cheap enough that they don't have a lot of price flexibility. If you tried to squeeze $20/mo ChatGPT subscriptions up to $200/mo then customers would disappear to other providers.
- And if you instantly switch off model building, 90% of the demand for datacenters evaporates, NVDA collapses, the Neoclouds crater to zero and go bankrupt -- AND THIS IS THE AI BUBBLE COLLAPSING.
This is one thing I find more on Hacker News than most other discussion sites. Any topic about annoying or bad behaviour by corporations eventually devolves to "they're allowed to do it".
"They're allowed to do it" is one of those things that doesn't really bear saying so much because that's already been implicitly established. Nobody is claiming what they're doing is illegal, so how does this add anything of value? Who is still unaware that corporations will use their size and market control to their benefit?
I'd really like to turn off the feature that tries to correct some previous word based on the next word. Most often, that mangles the last word because I swyped the next word wrong, or I just really wanted to type something different.
There are several levels to the autocorrection, so you can definitely do this. I've turned most autocorrect off on my iPhone at this point, because the people I talk to on a regular basis have a much easier time figuring out what I meant from regular misspellings than from the complete rewrite of phrases that iOS is fond of doing.
The inflation being caused by all the "crowding-out" caused by AI overinvestment is currently incredibly harmful.
That really needs to unwind.
However much potential energy has been stored in it, is going to turn into kinetic energy in the subsequent correction/crash.
The best time for this to happen is as soon as possible, causing the least amount of damage.
I pretty much expect the damage to be immense, though, but that is going to be unavoidable at this point. However big the bubble is, it needs to pop.
In retrospect, we will find that there were truly idiotic decisions made in blowing up the bubble, and in the aftermath the wrong people will almost certainly get hurt, because that is just what we do these days.
Advertisement and propaganda are two different arms of the same mechanism/technology stack.
And they have been for a very long time, dating back to at least when von Braun was building those V2s.
If we lived in a more humane world, intrusive advertisement and propaganda would all be banned--but that is like talking about some kind of utopia-level civilization right now. Might as well be discussing modern democracy back in the 10th century.
Maybe mathematics needs to end the "credit" for whoever discovered something first.
If you vomited out a bunch of AI slop that nobody read, history shouldn't care.
The first person to package it up for human consumption and get everyone to agree it is correct should get all the credit now. If you can't get up in front of a seminar and have pretty much everyone walk out understanding what you meant... then too bad, history forgets who you were...
From the evidence, one might dare to suggest that public education in the Soviet Union was pretty good, particularly of mathematics and the sciences. For example, the Math Circle as a social activity.
I'm also unconvinced that "Hard==Better" and that aggressive flunk-out programs are the best idea for a society.
As someone who hated undergrad math and physics approaches to vector calculus--until I found _Div, Grad, Curl and All That_--I'd argue that shitty hard textbooks are just shitty. There seems to be a cultural bias toward asceticism here, where suffering purifies the soul.
"Anthropic's Claude AI has quietly started restricting the kinds of work it will do, including modding, decompiling, or reverse-engineering video game code."
Uh, we're not allowed to use Claude to even mod games that fully support mods?
Based on how every time I ask it a virology question about SARS-CoV-2 and I get a refusal, I'm quite worried about how far-reaching this will be.
Anthropic is probably the worst company to look at because they lease all their datacenter space. They don't carry the debt, they don't play the games with depreciation, and they're probably most likely to survive because of that. Detailed analysis of Anthropic's S-1 in isolation tells you almost nothing about the problematic areas of the AI sector.
And there are a few really comical statements that render this whole thing unserious:
> The reason why AI companies are spending so much money is to rollout data centres to run the AI in the first place. These expenses are one-off purchases, not regular operating expenses. Once all the data centres are built, AI spending is going to plummet.
Okay, shares of NVDA just crashed to a round-off error close to zero.
There are other reasons why that statement is not accurate, but clearly the author hasn't even bothered to think through their argument at all.
> In a pinch, AI companies can also cut costs by no longer allowing people to use their products for free and temporarily suspending expensive AI research. This can be implemented instantly, whereas other kinds of expenses like leases and spending commitments are a lot harder to cut.
- AI companies are limited in their ability to jack up prices by open models being good enough and inference being cheap enough that they don't have a lot of price flexibility. If you tried to squeeze $20/mo ChatGPT subscriptions up to $200/mo then customers would disappear to other providers.
- And if you instantly switch off model building, 90% of the demand for datacenters evaporates, NVDA collapses, the Neoclouds crater to zero and go bankrupt -- AND THIS IS THE AI BUBBLE COLLAPSING.
reply