Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Pause Giant AI Experiments: An Open Letter

futureoflife.org

Post reply on HN
Print view View original

301–310 of 1001 posts

Re: Pause Giant AI Experiments: An Open Letter

#301

Post by labrador »

tgv wrote: post #281
labrador wrote: post #245

Elon thought AI would enslave us, but turns out it empowers us. Now he wants to shut it down and have it all to himself. It's getting to be that whatever the opposite of what Musk wants is the correct course of action for the people.

> Elon thought AI would enslave us, but turns out it empowers us. I'm not sure how you reached that conclusion. GPT3 is 2 years old. It has ample time to enslave us. I'm also still waiting for the "empowering" bit. Haven't really noticed it, I must say. Or are you referring to a few code completion tricks?

If you don't feel empowered by AI, imho you're doing it wrong and will probably be one of those left behind to be enslaved. I'm confident in my ability to stay on top, with a help from my own AI.

Re: Pause Giant AI Experiments: An Open Letter

#302

Post by thomastjeffery »

The dumb criticize the blind.

What an absurd situation! How did we get here? Here are the steps:

1. Large Language Models have been presented as "AI", which personifies them instead of describing how they work.

2. Goals for LLM development were set for the personified attributes, and not the actual functionally of the real thing. OpenAI brags about how GPT4 scores at human tests: as if that has any bearing on the model itself, and not simply its content.

3. The success of an LLM at a personified goal is determined by magic. Does the black box output what I am hoping it will? How do I improve it? Weights? Better training corpus?

4. An LLM is understood to be a person in a black box, and we could not possibly understand what that person will do next. Danger!

Re: Pause Giant AI Experiments: An Open Letter

#303

Post by permo-w »

tgv wrote: post #281
labrador wrote: post #245

Elon thought AI would enslave us, but turns out it empowers us. Now he wants to shut it down and have it all to himself. It's getting to be that whatever the opposite of what Musk wants is the correct course of action for the people.

> Elon thought AI would enslave us, but turns out it empowers us. I'm not sure how you reached that conclusion. GPT3 is 2 years old. It has ample time to enslave us. I'm also still waiting for the "empowering" bit. Haven't really noticed it, I must say. Or are you referring to a few code completion tricks?

the comment you're replying to is pretty delusional to say the least, but I disagree that they aren't empowering now. ChatGPT is an extremely useful source of education that bypasses the mess that is Google, and it's much more than just code completion tricks. gpt-4 can literally write long, complex programs that generally work the first time you run them

Re: Pause Giant AI Experiments: An Open Letter

#304

Post by dragonwriter »

labrador wrote: post #245

Elon thought AI would enslave us, but turns out it empowers us. Now he wants to shut it down and have it all to himself. It's getting to be that whatever the opposite of what Musk wants is the correct course of action for the people.

> Elon thought AI would enslave us, but turns out it empowers us.

Absent other non-technical changes, AI along the lines that the main commercial proponents are pursuing empowers the narrow set of capitalists who control it to more efficiently extract value from and exert power over everyone else.

“Empower us” and “enslave us” are both accurate, for different values of “us”.

(A few people riding the top of the capitalist-control-of-AI wave are trying to sell it as an existential threat to everyone, and to use that as an excuse to concentrate control of AI in their hands, to make social change in the opposite direction needed to mitigate the threat they themselves pose with AI; is this a real threat? If the shared threat is real, it is nowhere near as imminent and probable as the threat posed by those who are selling it.)

Re: Pause Giant AI Experiments: An Open Letter

#305

Post by permo-w »

labrador wrote: post #301
tgv wrote: post #281

Earlier quoted context omitted.

> Elon thought AI would enslave us, but turns out it empowers us. I'm not sure how you reached that conclusion. GPT3 is 2 years old. It has ample time to enslave us. I'm also still waiting for the "empowering" bit. Haven't really noticed it, I must say. Or are you referring to a few code completion tricks?

If you don't feel empowered by AI, imho you're doing it wrong and will probably be one of those left behind to be enslaved. I'm confident in my ability to stay on top, with a help from my own AI.

so when some societal reject with a CS degree sets up an intention bot to hack into/socially engineer the release of more gm viruses - or worse - what's your personal AI going to do about that?

this isn't about being enslaved, it's about being hurt by the actions of a human-like you-who thinks that AI is their route to the top, or just wants to watch the world burn. AI enslavement makes for good films, but we'll all be long dead before that happens

Re: Pause Giant AI Experiments: An Open Letter

#306

Post by prohobo »

thatguy0900 wrote: post #276
dylan604 wrote: post #274

Earlier quoted context omitted.

If I don't do it, someone else will, so it might as well be me. At least I know the morals I have /s

This but without the s. What does the world look like if only China and Russia have agi because the western world took a pause to consider the ethics. This tech is the same as nukes in that sense.

Maybe a world where China and Russia might need our ethical controls to tame the AI they let loose on themselves?

At this point I feel like we deserve to lose, our people are insane and can only operate at a shallow level of game theory that ensures everyone behaves like teenagers.

That said, I'm not sold on needing to pause. We're likely hitting the limit of LLMs for the next while - only incremental improvements from here. What's wrong with creating a task force and committee for working on AI ethics in parallel?

Re: Pause Giant AI Experiments: An Open Letter

#308

Post by permo-w »

thomastjeffery wrote: post #302

The dumb criticize the blind. What an absurd situation! How did we get here? Here are the steps: 1. Large Language Models have been presented as "AI", which personifies them instead of describing how they work. 2. Goals for LLM development were set for the personified attributes, and not the actual functionally of the real thing. OpenAI brags about how GPT4 scores at human tests: as if that has any bearing on the mod…

I'm not concerned about how they work or whether they're attaining personhood or not. I'm concerned at what people can and will use them for to the detriment of our civilisation

Re: Pause Giant AI Experiments: An Open Letter

#310

Post by adamsmith143 »

walkhour wrote: post #251

Breaking: the runners lagging behind in the AI race demand openAI to stop right on its tracks; this is very unsafe; it will be much safer once we catch up and openAI is on the rear view mirror.

This is frankly a bad and cavalier take on an extremely important subject. Many on the list are Academics outside AI/ML and/or leaders of AI orgs at the very top of the field that have no need to catch up to or slow down OpenAI to benefit themselves. Risks from AI are very real and Sam Altman himself has said so numerous times. He also in fact advocated for slowing down AI progress on Lex Fridman's podcast this month.
Post reply on HN