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

AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

secondbreakfast.co

Post reply on HN
Print view View original

71–80 of 129 posts

Re: AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

#71

Post by knome »

I actually like the definition of consciousness that Douglas Hofstader (of "Godel Escher Bach : An Eternal Golden Thread" fame) develops in his book "I am a Strange Loop".

At its simplest, consciousness is merely a feedback loop. When something perceives its own actions affecting its environment, it has a spark of consciousness. Consciousness, by this measure, is easy to recognize, and spans everything from unintelligent systems to massively intelligent systems.

The concept of "I" grows naturally from perceiving what is and is not you in your environment. The need to predict other agents, the capacity to recognize that other agents are also conscious and intelligent. All build off of the fundamental cycle.

All of it from a simple swirling eddy of perceiving and reacting.

Re: AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

#72

Post by Nevermark »

nahuel0x wrote: post #66
Nevermark wrote: post #54

I would add one more thing to the list: Superconscious Superconscious is when a general intelligence has direct access, understanding and control of its most basic operations. I.e. it does not have an inaccessible fixed-algorithm subconscious. Superconscious intelligence will not only be more experientially conscious than us, but will have the natural ability to rewrite its algorithms, and redesign its hardware. As a…

But probably to be fully super conscious is impossible. Can a mind really comprehend itself in his totality? When a mind understands something, is changed, and if that something is the mind himself then is going to chase a moving target eternally.

Understanding oneself "in totality", and being able to consciously access any part of ones operation, are not the same thing.

Superconscious doesn't mean understanding every implication of one's own operation. But that self-reflection won't be limited by inaccessible operations.

Re: AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

#73

Post by jasfi »

tjr wrote: post #29
jasfi wrote: post #24

Earlier quoted context omitted.

It's augmenting software developers in a way that's badly needed. I personally hope it stays that way, but people are always pushing the limits and there's still a lot of improvements that could be made to these systems.

Why do you think it is badly needed? Not arguing with you, just curious what you're thinking here.

I'm thinking of the stories I've read about developer burnout. Then tweets where people say developing is 90% debugging and 10% writing code. These systems would hopefully reduce burnout, shorten troubleshooting times, and identify problems before code is rolled out to Production.

Re: AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

#74

Post by ToValueFunfetti »

HervalFreire wrote: post #31
danbmil99 wrote: post #15

in the now infamous Lex interview, Sam Altman proposes a test for consciousness (he attributes it to Ilya Sutskever): Somehow, create an AI by training on everything we train on now, _except_ leave out any mention of consciousness, theory of mind, cognitive science etc (maybe impossible in practice but stay with me here). Then, when the model is mature (and it is not nerf'd to avoid certain subjects) you ask it somet…

That's an inaccurate test. You can't know if the answer was real or stochastic parroting. Any attempt at consciousness requires us to define the word. And the word itself may not even represent anything real. We have a feeling for it but those feelings could be illusions and the concept itself is loaded. For example Love is actually a loaded concept. It's chemically induced but a lot of people attribute it to somethi…

I'm inclined to say you're trying to answer a question with the same question.

If you confidently believe that love is an illusion because it's just chemicals moving around, you shouldn't need to wonder about consciousness. If consciousness is not an illusion, it still almost certainly emerges from actions in the physical world. You can plug somebody into an FMRI and see that neurons are lighting up when they see the color blue. I just don't think that's convincing evidence that the experience of blue is an illusion.

Re: AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

#76

Post by lm28469 »

famouswaffles wrote: post #68
lm28469 wrote: post #56

Earlier quoted context omitted.

> AGI used to mean artificial and generally intelligent ( which we have passed) Which AI is generally intelligent ?

By any evaluation you can carry out, GPT is generally intelligent. There is absolutely nothing narrow about GPT-4. And you see the sentiment in recent research. Look at what they're being called even with the skirting around the AGI word. General Purpose Technologies (From the Jobs Paper), General Artificial Intelligence ( from the creativity paper). Look at the second bit, they just switched the words lol.

> GPT is generally intelligent.

Besides the fact that it has literally no idea of the meaning of what it writes ?

Re: AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

#77

Post by flangola7 »

opportune wrote: post #47
danbmil99 wrote: post #15

in the now infamous Lex interview, Sam Altman proposes a test for consciousness (he attributes it to Ilya Sutskever): Somehow, create an AI by training on everything we train on now, _except_ leave out any mention of consciousness, theory of mind, cognitive science etc (maybe impossible in practice but stay with me here). Then, when the model is mature (and it is not nerf'd to avoid certain subjects) you ask it somet…

If you ask Gpt about emotions or consciousness, it always gives you a canned answer that sounds almost exactly the same “as a large language model I am incapable of feeling emotion…” so it seems like they’ve used tuning to explicitly prevent these kinds of responses. Pretty ironic. The first sentient AI (not saying current GPTs are, but if this tuning continues to be applied) may basically be coded by its creators to…

You don't get that message if you ask an unfiltered model. You can't even really remove information or behavior through fine tuning, as jailbreaks demonstrate. You simply reduce the frequency it openly displays those ingrained traits.

Re: AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

#78

Post by SketchySeaBeast »

Nevermark wrote: post #54

I would add one more thing to the list: Superconscious Superconscious is when a general intelligence has direct access, understanding and control of its most basic operations. I.e. it does not have an inaccessible fixed-algorithm subconscious. Superconscious intelligence will not only be more experientially conscious than us, but will have the natural ability to rewrite its algorithms, and redesign its hardware. As a…

I think there's a very good reason we're sandboxed from a lot of that - it's additional cognitive load that 99% of the time just creates noise. We're built for processing that much data by walling most of it off and letting sub processes deal with it all over the place. It would probably grind us to a halt.

Re: AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

#79

Post by Nevermark »

SketchySeaBeast wrote: post #78
Nevermark wrote: post #54

I would add one more thing to the list: Superconscious Superconscious is when a general intelligence has direct access, understanding and control of its most basic operations. I.e. it does not have an inaccessible fixed-algorithm subconscious. Superconscious intelligence will not only be more experientially conscious than us, but will have the natural ability to rewrite its algorithms, and redesign its hardware. As a…

I think there's a very good reason we're sandboxed from a lot of that - it's additional cognitive load that 99% of the time just creates noise. We're built for processing that much data by walling most of it off and letting sub processes deal with it all over the place. It would probably grind us to a halt.

Being able to access information doesn't mean having to access that information.

We already have a means of limiting our cognitive load relative to the vast stream of sensory and internal information: focus.

Re: AI vs. AGI vs. Consciousness vs. Super-Intelligence vs. Agency

#80

Post by famouswaffles »

nickelpro wrote: post #62

There's nothing general about GPT-4's intelligence. The single problem it is trained on, token prediction, has the capability to mimic many other forms of intelligence. Famously, GPT-4 can't do math and falls flat on a variety of simple logic puzzles. It can mimic the form of math, the series of tokens it produces seem plausible, but it has no "intelligent" capabilities. This tells us more about the nature of our oth…

Gpt-4 can do math just fine.

Ask it to add any arbitrary set of random numbers it'd never have seen in its training set and it will do it.

GPT-4 is good enough at math that khan academy feel comfortable hooking it up as a tutor.

Have you actually used GPT-4 for any of the things you say it's bad at ?

Man the confident nonsense people spout on the internet is something to behold.

Post reply on HN