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

Why This AI Moment May Be the Real Deal

thenewatlantis.com

Post reply on HN
Print view View original

71–80 of 231 posts

Re: Why This AI Moment May Be the Real Deal

#71

Post by guhidalg »

ZiiS wrote: post #37
roenxi wrote: post #2

> Study Deep Blue’s 1997 defeat of world chess champion Garry Kasparov, and your excitement about how quickly this technology would take over other cognitive work would wane as you learned just how much brute human force went into fine-tuning the software specifically to beat Kasparov. That was ~26 yeas ago. In the intervening 26 years humans have been curb-stomped so hard at chess that certain microwaves might be co…

The are microwaves fully in on the curb-stomping https://www.tomsguide.com/news/ges-kitchen-hub-is-the-smarte... My back of the envelope suggests most digital microwaves with a smart defrost but no connectivity are probably more in the 2000 ELO range (i.e. can beat most players but not grandmasters). Mainly due to not having enough storage for large opening/endgame books.

I for one welcome measuring compute power in chess ELO over FLOPS. Let's me know which devices I can trounce and which ones to pay respect to.

Re: Why This AI Moment May Be the Real Deal

#72

Post by nyolfen »

jjoonathan wrote: post #49
Plyphon_ wrote: post #43

Earlier quoted context omitted.

It's just a turn of phrase that means you got 100% beaten without challenge.

No, no it isn't. It's a very particular violent act designed to maximize ongoing physical pain and social disadvantage in the victim. Sort of like throwing sulfuric acid at someone's face.

wait until you hear such idioms as "punched in the gut", "made a killing", etc.

Re: Why This AI Moment May Be the Real Deal

#74

Post by flangola7 »

danaris wrote: post #52
gmerc wrote: post #48

Earlier quoted context omitted.

Not really, hook up external sensors to keep shoveling data , feedback data into continuous lora and performance comparison,

How? What kind of data? You think you can just hook up a firehose of visual or auditory data to ChatGPT and have that produce anything meaningful? That's not even the basic format that ChatGPT operates in. Furthermore, isn't part of the point of the dataset an LLM is trained on that it has to be at least moderately structured and tagged, or you end up with garbage in the output?

Meta has an LLM trained on multiple senses, including images, sound, video, physical motion, heat, and lidar depth.

Re: Why This AI Moment May Be the Real Deal

#75

Post by unshavedyak »

strikelaserclaw wrote: post #67
mercurialsolo wrote: post #62

It's not combining an LLM with data/memory recollection and infinite agency that will lead to leaps in cognition. Complex systems still can do that and still not have self-awareness / cognition. We will need to take some more fundamental leaps to synthesize memory, agency and prediction to operate within a single system where self-awareness resides in the entity and not in the building blocks. LLM's represent a build…

Although i fundamentally agree with you, i will say that it is impossible to know which "tricks" will lead to a leap in cognition, even looking evolutionarily, the difference in genome between a monkey and human is very small percentage wise, yet the effects of that small difference are stark.

I'm always on the position that at some point (i make no prediction of when) "AI" will still be labeled as not-intelligent, but to external observers it won't be possible to know.

Ie we will be able to describe how a function of `add(1,2)` is most clearly definitely not intelligent. We will understand the internal makeup of these neurons better, the whole system will be deterministic and understood - just a very large `add(1,2)` .. so it's definitely not intelligent, of course.

Yet it will still produce output that is fully and truly turing-proof. If you let it run, it retains, stores, "behaves", etc. It's clearly just `add(1,2)` in a loop after all.. we will understand it so well that it's obviously just a trick of our mind that it seems real.

The conclusion in that space will not be one of the machine .. but of us. If we can't tell it apart from us.. and if we know it is most definitely not alive or even mysterious.. what does that say about us? The existential crisis will be neat.

Re: Why This AI Moment May Be the Real Deal

#76

Post by hackinthebochs »

danaris wrote: post #51
hackinthebochs wrote: post #45

Earlier quoted context omitted.

Why?

Well, "consciousness", at least as we recognize it, requires a mechanism by which the entity being measured can continuously form new "thoughts" and "memories" (which requires continuity of learning, and at the very least continuity of input being fed back from its own output), and some form of continuous external input of information about the world to at least be available , even if it is not always on. A standard…

Continuous processing may be putting undue weight on accidental features of humans to be a necessary feature. Consider that a conscious mind can't represent the gaps in its processing, and so has an appearance of continuity. But this appearance probably doesn't map onto a continuous reality. An example is anesthesia patients that report a seemingly uninterrupted time from counting down pre-surgery to waking up. So interruptions, gaps, discontinuities, and so on don't necessarily eliminate the possibility for consciousness. It may be the case that LLMs are conscious when they are engaged in active inference.

While I generally favor a requirement for recurrent processing, I have low but non-zero credence for certain feedforward networks being conscious. The point of recurrence is to allow information about itself to influence its processing. But It seems plausible that feedforward constructs can represent meta-information in a way that is computed as part of constructing the output.

Re: Why This AI Moment May Be the Real Deal

#77

Post by ImHereToVote »

mercurialsolo wrote: post #62

It's not combining an LLM with data/memory recollection and infinite agency that will lead to leaps in cognition. Complex systems still can do that and still not have self-awareness / cognition. We will need to take some more fundamental leaps to synthesize memory, agency and prediction to operate within a single system where self-awareness resides in the entity and not in the building blocks. LLM's represent a build…

Why not use less emotionally charged words? Goal to action mapping for instance. How far away are we from a general goal to action mapper that can outcompete a human being on a given goal.

Re: Why This AI Moment May Be the Real Deal

#78

Post by adrianN »

danaris wrote: post #51
hackinthebochs wrote: post #45

Earlier quoted context omitted.

Why?

Well, "consciousness", at least as we recognize it, requires a mechanism by which the entity being measured can continuously form new "thoughts" and "memories" (which requires continuity of learning, and at the very least continuity of input being fed back from its own output), and some form of continuous external input of information about the world to at least be available , even if it is not always on. A standard…

We don’t have a definition of consciousness that allows you to recognize it.

Re: Why This AI Moment May Be the Real Deal

#79

Post by c_crank »

unshavedyak wrote: post #75
strikelaserclaw wrote: post #67

Earlier quoted context omitted.

Although i fundamentally agree with you, i will say that it is impossible to know which "tricks" will lead to a leap in cognition, even looking evolutionarily, the difference in genome between a monkey and human is very small percentage wise, yet the effects of that small difference are stark.

I'm always on the position that at some point (i make no prediction of when) "AI" will still be labeled as not-intelligent, but to external observers it won't be possible to know. Ie we will be able to describe how a function of `add(1,2)` is most clearly definitely not intelligent. We will understand the internal makeup of these neurons better, the whole system will be deterministic and understood - just a very larg…

>If we can't tell it apart from us.. and if we know it is most definitely not alive or even mysterious.. what does that say about us?

Less than people think. We know that being a biological system differentiates us. If intelligent systems tend to converge where language is concerned, so be it. It doesn't mean the AI warrants any special consideration. It's still a tool.

Re: Why This AI Moment May Be the Real Deal

#80

Post by balnaphone »

Am I misreading this?

> How tall was the president when JFK was born?

I infer that "JFK" is being referred to as "the president", regardless of what stage of life he's at. Therefore the correct answer would be the length of his body as a newborn, with nothing to do with who was the president at the moment of his birth.

For example, given the headline: "Was the president strong at writing essays when JFK was a high school freshman? Read this essay to find out!", would you expect an essay from Herbert Hoover while he was in office, or an essay from JFK when he was in grade 9?

Post reply on HN