Live data from Hacker News

Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

businessinsider.com

21–30 of 873 posts

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#21
That's the thing with AI. It's so dangerous it could destroy us all. P(doom)=0.3.

Anyway, that's why we're working on much large LLMs. It's important that we do this research and expand the size of the LLMs because a larger LLM could destroy us all.

Listen, everyone cosplays as doomers here because it's stylish right now. But they're all trying to make this thing they think is suicide. Is this a suicide cult?

No. It's just people playing around acting important while solidifying their business.

Recently, I read on Twitter that many AI CTOs think that P(doom) > 0.25 by 2030. Here's my question to one CTO of a venture backed firm with present personal assets over $400k that believes this:

- A representative of mine will meet you in South Park

- The representative will be carrying a $100k cheque and a document outlining terms for you to pay a lump sum of $177k on Feb 1 2030 (10% compounding, roughly)

- The loan will be unsecured

- We will make the terms public

This is pretty close to break-even for you at terms you won't receive personally anywhere else. Once doom happens money will be worthless. Take my money. I'll also make the deal all numbers divided by 100 if you just want to have some fun.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#22
post #20

The moat he describes the one-percenters might aquire is government regulation. Creating an environment where only an exclusive club of companies is able to work on AI. Is that possible? Has that happened in any other industry? I can imagine it happening in a single country. But wouldn't research route around that country then? How is the situation in Europe, China, India, Canada, Japan ... ?

Ehhh capital alone is plenty of a moat, esp when there’s no regulation to keep the market fair.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#24
post #20

The moat he describes the one-percenters might aquire is government regulation. Creating an environment where only an exclusive club of companies is able to work on AI. Is that possible? Has that happened in any other industry? I can imagine it happening in a single country. But wouldn't research route around that country then? How is the situation in Europe, China, India, Canada, Japan ... ?

Yes, the US government could try and do what they did with cryptography in the 90s and succeed.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#25
post #6

Earlier quoted context omitted.

> It's too late for monopolizing large language models. The technology for making them is known and the cost keeps coming down. Not so sure non-institutional folks can catch up. OpenAI is trying to regulatory capture LLMs to prevent the Amazons and the well-funded unicorns from following. Everyone left in the lurch after these players crystalize is SOL. You, average HN reader, cannot possibly compete with OpenAI. The…

Amazon excels at regulatory compliance. This isn’t a moat against Amazon. It’s a moat against you.

The same is true of Microsoft, who more or less own OpenAI now.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#27
post #20

The moat he describes the one-percenters might aquire is government regulation. Creating an environment where only an exclusive club of companies is able to work on AI. Is that possible? Has that happened in any other industry? I can imagine it happening in a single country. But wouldn't research route around that country then? How is the situation in Europe, China, India, Canada, Japan ... ?

>But wouldn't research route around that country then?

it already does. Stable Diffusion originated at LMU Munich with a lot of non-profit funding, EleutherAI and LAION are non-profit as well I think. The Chinese companies all have their own models anyway etc.

So yes I also find it unlikely. Of course there'll always be one or two players a little bit ahead of the field at any given time but this stuff is almost entirely public research based. These models largely run on a few thousand lines of Python, public data and a ton of compute. The latter is the biggest bottleneck but that'll diminish as more manufacturers come online.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#28
post #12
post #2

some speculators say we'll enter an age of abundance and money wont matter nearly as much (or at all) but if thats the case, how do businesses transition from their current state of profit over everything into money is not important? I see it as much more likely that things become monopolized instead

Eh, the strong market for virtual goods and the existence of significant virtual economies[1] like those found in Roblox and Second Life lead me to believe that markets can and will always exist, regardless of how far we progress into and beyond post-scarcity (assuming humanity persists in its current form and consumption preferences remain stable). I'm less certain about the capitalist model in particular, but it to…

We are already in a post-scarcity world, we have been for decades.

The problem is inequalities and the inefficiency of how we use the resources, it doesn't look like we'll change that soon, sadly. I'm more prone to think that in a "virtual" economy we'll still have miserable workers to empty the chamber-pots of the lucky ones who are plugged in the new world.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#29
post #4

It's too late for monopolizing large language models. The technology for making them is known and the cost keeps coming down. There are things to worry about, but this doesn't seem to be one of them.

Training is still very expensive.. and if it becomes cheaper, then a next-gen model with better performance will be expensive again. It may be very hard to train cutting edge models if you only have consumer hardware

Why do you need cutting edge models if you can solve your problems with less?

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#30
post #4

It's too late for monopolizing large language models. The technology for making them is known and the cost keeps coming down. There are things to worry about, but this doesn't seem to be one of them.

Cost of technology doesn't matter for this.

It's not the hammers, it's the building permits.

Post reply on HN