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

OpenAI Announces $10M Superalignment Grants

openai.com

Post reply on HN
Print view View original

51–60 of 61 posts

Re: OpenAI Announces $10M Superalignment Grants

#51

Post by p1esk »

adastra22 wrote: post #41
p1esk wrote: post #4

Earlier quoted context omitted.

I think it's pretty clear what "superintelligence" means, no? Superintelligence: smarter than any human. AGI: as smart as an average human. "As smart" means it can do anything a human can do.

No, it’s not clear. There are already many domains where AI is vastly smarter than any human. There are algorithms like GPT which are fully general. So do we already have superhuman AGI today? Most people would object. But it is hard to argue against that on definitions. It usually ends up being “by superintelligence I meant the singularity.” Which really makes this a question about reality matching science fiction…

Smarter than any human in every domain. So no, we don’t have a superhuman AGI yet. And it has nothing to do with singularity.

Re: OpenAI Announces $10M Superalignment Grants

#52

Post by famouswaffles »

eastof wrote: post #13
p1esk wrote: post #4

Earlier quoted context omitted.

I think it's pretty clear what "superintelligence" means, no? Superintelligence: smarter than any human. AGI: as smart as an average human. "As smart" means it can do anything a human can do.

Aren't those terms totally meaningless in terms of modern "AI" though? AGI is nowhere near being able to play a violin solo, their fine motor skills are abysmal. The list goes on and on, they also aren't running a bakery, collecting field data for ecological studies, or playing literally any sport any time soon either. The concept of a disembodied system even being "smart" doesn't make sense to me.

There are literally millions, millions of humans unable to do any of the things you have listed. Do they no longer qualify as general intelligences ?

Re: OpenAI Announces $10M Superalignment Grants

#53

Post by famouswaffles »

BobaFloutist wrote: post #26
p1esk wrote: post #7

Earlier quoted context omitted.

Yes, when put in a robotic body.

Ok so that's no "AI" that currently exists then.

>Ok so that's no "AI" that currently exists then.

Lol https://www.youtube.com/watch?v=CnkM0AecxYA

This is far from the first of this class of LLM robots too https://tidybot.cs.princeton.edu/

Re: OpenAI Announces $10M Superalignment Grants

#54

Post by famouswaffles »

dumbo-octopus wrote: post #50
selfhoster11 wrote: post #44

Earlier quoted context omitted.

No. Is a human body that suffers from selected kinds of sickness for that matter? Also no.

Pardon? Are you saying that because a human can't heal from all kinds of sickness, healing from some kinds isn't "a thing a human can do"?

Not every human body is really capable of any of "metabolism, growth, self repair"

Do you want a list of the kinds of afflictions that stunt these ?

For the ones that can't, they don't deserve to be called intelligent ?

What any of that has to do with intelligence is honestly beyond me.

Re: OpenAI Announces $10M Superalignment Grants

#55

Post by dumbo-octopus »

famouswaffles wrote: post #54
dumbo-octopus wrote: post #50

Earlier quoted context omitted.

Pardon? Are you saying that because a human can't heal from all kinds of sickness, healing from some kinds isn't "a thing a human can do"?

Not every human body is really capable of any of "metabolism, growth, self repair" Do you want a list of the kinds of afflictions that stunt these ? For the ones that can't, they don't deserve to be called intelligent ? What any of that has to do with intelligence is honestly beyond me.

> Not every human body is really capable of any of "metabolism, growth, self repair"

99.9% percent of them are, and those that aren't would be considered malformed.

> For the ones that can't, they don't deserve to be called intelligent? What any of that has to do with intelligence is honestly beyond me.

You haven't been paying attention to this thread, yet you reply to it? Please reread it, in particular the parent's argument I was responding to. Such context would be helpful to have when responding to me, yes?

Re: OpenAI Announces $10M Superalignment Grants

#56

Post by adastra22 »

p1esk wrote: post #51
adastra22 wrote: post #41

Earlier quoted context omitted.

No, it’s not clear. There are already many domains where AI is vastly smarter than any human. There are algorithms like GPT which are fully general. So do we already have superhuman AGI today? Most people would object. But it is hard to argue against that on definitions. It usually ends up being “by superintelligence I meant the singularity.” Which really makes this a question about reality matching science fiction…

Smarter than any human in every domain. So no, we don’t have a superhuman AGI yet. And it has nothing to do with singularity.

Write a script which downloads huggingface models and tries each in turn. Voila, smarter than human in every domain.

Ah, but “it can’t fold laundry” you say. When was the last time you heard someone non-ironically describe folding laundry as something requiring human-level intelligence?

You are hiding behind the ill defined term “smart” and performing the classic motte and bailey maneuver. If you honestly attempt to match “intelligence” or “smarter than” to its colloquial meaning of “being able to understand things, apply learned knowledge, and solve problems that require thinking” then yes, we have super intelligent AIs, and have had for many years.

Re: OpenAI Announces $10M Superalignment Grants

#57

Post by p1esk »

adastra22 wrote: post #56
p1esk wrote: post #51

Earlier quoted context omitted.

Smarter than any human in every domain. So no, we don’t have a superhuman AGI yet. And it has nothing to do with singularity.

Write a script which downloads huggingface models and tries each in turn. Voila, smarter than human in every domain. Ah, but “it can’t fold laundry” you say. When was the last time you heard someone non-ironically describe folding laundry as something requiring human-level intelligence? You are hiding behind the ill defined term “smart” and performing the classic motte and bailey maneuver. If you honestly attempt to…

What model should I download that would do my work for me today? I’ll explain the project I’m working on, and will give it full access to the codebase, documentation, Slack, and Zoom so it can find all the context and ask my colleagues and my manager if it needs more info. GPT-4 is obviously not smart enough to do this (I’ve been using it daily since it came out), but it seems like you know of some miracle models on HF which can? Would love to try them out.

Re: OpenAI Announces $10M Superalignment Grants

#58

Post by adastra22 »

p1esk wrote: post #57
adastra22 wrote: post #56

Earlier quoted context omitted.

Write a script which downloads huggingface models and tries each in turn. Voila, smarter than human in every domain. Ah, but “it can’t fold laundry” you say. When was the last time you heard someone non-ironically describe folding laundry as something requiring human-level intelligence? You are hiding behind the ill defined term “smart” and performing the classic motte and bailey maneuver. If you honestly attempt to…

What model should I download that would do my work for me today? I’ll explain the project I’m working on, and will give it full access to the codebase, documentation, Slack, and Zoom so it can find all the context and ask my colleagues and my manager if it needs more info. GPT-4 is obviously not smart enough to do this (I’ve been using it daily since it came out), but it seems like you know of some miracle models on…

Why is GPT not smart enough to do this? That is not obvious to me.

Re: OpenAI Announces $10M Superalignment Grants

#59

Post by p1esk »

adastra22 wrote: post #58
p1esk wrote: post #57

Earlier quoted context omitted.

What model should I download that would do my work for me today? I’ll explain the project I’m working on, and will give it full access to the codebase, documentation, Slack, and Zoom so it can find all the context and ask my colleagues and my manager if it needs more info. GPT-4 is obviously not smart enough to do this (I’ve been using it daily since it came out), but it seems like you know of some miracle models on…

Why is GPT not smart enough to do this? That is not obvious to me.

Then you should try it, it will become obvious pretty fast.

Re: OpenAI Announces $10M Superalignment Grants

#60

Post by righthand »

selfhoster11 wrote: post #48
righthand wrote: post #36

Earlier quoted context omitted.

It tastes like puppy chow. You’re comfortable eating the same thing every day for every meal right?

Why the heck not. If it saves me cooking and they rotate a couple of varieties so I don't get sick, I don't see a downside.

No varieties, variety is not optimal in feeding all humans. You do not purchase human chow at a store it is fed to you by the computers.
Post reply on HN