5 ms·
>> But despite encouraging results, most of the time I’m reminded that we’re nowhere near achieving human-like AI. Why? Primarily, because we have no idea what
by YeGoblynQueenne 7y ago
>> But despite encouraging results, most of the time I’m reminded that we’re nowhere near achieving human-like AI. Why?
Primarily, because we have no idea what intelligence is, or how it works, why it exists even, etc etc. This goes for human-like intelligence, but also for any kind of intelligence. We just have no good scientific understanding of the subject. We have some vague models of it ("the brain is like a computer and the mind is like a program running on it") but nothing very precise and certainly nothing that can be reproduced on a digital computer, which is what "human-like AI" would be (i.e. human-like AI would be the reproduction of human intelligence on a digital computer).
Most likely, until we make some progress in understanding intelligence we will not be able to reproduce it. Except perhaps by chance.
- noobiemcfoob 7y agoHow curious that one of the best models of intelligence is one of its products created in an effort to emulate itself.
- carapace 7y agoRight. Turing automated a part of the operation of his own mind, computers have always been AI. But it seems (to me) that logic doesn't capture all of what we call intelligence, and that the effort to recapitulate the human mind solely on the basis of discrete logic machines is almost certainly going to fail. But I think that failure will be valuable, and that we should not avoid the attempt to make a sculpture of the mind. Babbage had an automaton, the Silver Lady, that could imitate the dancing of a ballerina. We might be able to make "Silver Minds" that imitate the outward manifestations of the human personality, and so learn more about the nature and essence of thinking, as the Silver Lady must have of dancing. Certainly the craftsman who made her must have known a thing or two about ballet and physiology?
- olalonde 7y agoThat's largely the idea behind the concept of "technological singularity". We are quite slow at developing the first version (it's still a work in progress) but that v1 will likely be able to build a v2 much faster and so on, leading to an "intelligence explosion". [0] https://en.wikipedia.org/wiki/Technological_singularity https://en.wikipedia.org/wiki/Technological_singularity
- nradov 7y agoThere is no actual evidence that such a scenario is "likely". At this point it's just a bunch of unsupported assumptions and shaky analogies.
- vkou 7y agoThe inevitability of the Singularity, and most of the rest of the baggage train that comes with Kurzweilanism reads more like a religious conviction, and less like a scientific certainty. There are enormous gaps in reasoning that you just have to gloss over to go from the current state of the world to a post-singularity one, that you are expected to accept, on faith. I personally think it's about as likely as the return of the Messiah in our lifetimes. That is to say, it is not.
- sumitviii 7y agoKurzweilanism=Gavinism
- outworlder 7y ago> We are quite slow at developing the first version (it's still a work in progress) but that v1 will likely be able to build a v2 much faster and so on, leading to an "intelligence explosion". As with everything in nature, there are diminishing returns. It is not clear that any significant leaps in 'intelligence' are possible. At least to the extent proposed by 'singularity' advocates. Or that such an intelligence would be able to completely outclass humans, either in the current form, or augumented by tech.
- AgentME 7y agoAs humans have evolved larger brains, intelligence went up extremely. Human brain size is limited by a lot of factors specific to human biology, like the fact that it needs to be small enough to fit through the birth canal, it can't take up too much of the body's energy, and it can't make us too top-heavy. It seems unlikely that human evolution ran into these limits right at the same time it finished making a maximally-intelligent mind. None of these limiting factors apply to AI. If/when we figure out how to make intelligent software, we make AIs, and then we improve them to be human-level intelligent, it seems really unlikely to me that we'll hit a limiting factor right there. I think if we ever create near-human-level intelligent AIs, we'll blow right past human-level intelligence and hit a limit further on that's not coincidentally placed at human-level intelligence.
- eli_gottlieb 7y ago>Primarily, because we have no idea what intelligence is, or how it works, why it exists even, etc etc. This goes for human-like intelligence, but also for any kind of intelligence. We just have no good scientific understanding of the subject. Speaking as a machine learning and neuroscience researcher, how thoroughly have you investigated the matter to claim this? I would definitely not say that we have "no" understanding of the subject. We have a number of predominant paradigms and competing theories regarding the matter.
- dwaltrip 7y agoYou shouldn't be downvoted. That portion is clearly over-the-top hyperbole, and it is fair to call that out. We do have some understanding, even if it is somewhat limited.
- 50656E6973 7y agoHow would you define what intelligence is, in simple terms to a layman?
- eli_gottlieb 7y agoModel-based simulation and planning across a hierarchy of temporospatial scales. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3482298/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3482298/ https://www.sciencedirect.com/science/article/pii/S0959438813000172 https://www.sciencedirect.com/science/article/pii/S095943881... https://www.fil.ion.ucl.ac.uk/~karl/Whatever%20next.pdf https://www.fil.ion.ucl.ac.uk/~karl/Whatever%20next.pdf
- YeGoblynQueenne 7y agoWhat you describe sounds very much like a system that could be implemented with computers and software. Has anyone ever tried to do that?
- eli_gottlieb 7y ago
- trevyn 7y agoThis is incorrect. György Buzsáki‘s work is as fine a place as any to start.
- hammock 7y ago>Primarily, because we have no idea what intelligence is, or how it works, why it exists even, etc etc. I believe so many of us limit ourselves by even thinking that intelligence happens in the brain. By the title of this article I thought it might go into the vagus nerve, etc. There is so much "intelligence," processing, communication etc that happens away from the brain and independent of the brain in our bodies. We have to remember many animals (which we evolved from and retain vestiges of) don't have "centralized" intelligence and yet are capable of great things. We also have to remember that the trillions of cells we are made of each have their own nucleus as well.
- carapace 7y ago"What Bodies Think About: Bioelectric Computation Outside the Nervous System" https://news.ycombinator.com/item?id=18736698 https://news.ycombinator.com/item?id=18736698 Whatever it is, intelligence is ambient.
- danielam 7y ago> we have no idea what intelligence is I wouldn't make so strong a claim in light of what philosophy has to say about intelligence. Relevant here also are certain metaphysical presuppositions made by fields like neuroscience that preclude that understanding. Furthermore, even knowledge of something does not necessarily entail ability to reproduce something.
- dmreedy 7y agoPhilosophy absolutely has lots to say about intelligence, much of which is mutually contradicting. There are a lot of different models out there, based off of varying levels of cleavage to various priors, that all seem to have very little real explanatory (or, I should say, predictive) power. Approaches based off of current materialist, biology-based methods have a little more utility behind them as far as we can tell (some of our drugs seem to do something to the mind), but the furthest they've gotten so far is to be able to figure out some of the things that matter to intelligence, rather than a successful general model for what it is. And while I'll agree that this lack doesn't necessarily preclude the ability to reproduce intelligence, it does make it darn hard to recognize it.
- RosanaAnaDana 7y agoMaybe, or maybe not. Maybe much of what philosophy has to say to say about what it thinks about intelligence is bull-hooky and actually holding us back. The 'understanding' we have feeds into our presumptions, so I agree in that regard. Arguably, the tangible limits experienced in our actual models of the world based on those presumptions are reflective of the validity of those assumptions. Clearly, in the case of modeling 'intelligence' and 'mind', they aren't that great, indicating we really don't understand the phenomena. I've made this argument repeatedly and before on this platform, but I think the biggest mistake we've made in this field is regarding the preeminence of what we experience as 'human' intelligence and our mental models for how that 'must' work. I'm going to catch all kinds of flack for this, but its my biased perspective that plants are just as intelligent as animal life, but because we use humans as a litmus for what intelligence must 'look' like; we have a very difficult time enumerating what intelligence a non-human organism must have. I think intelligence is reliant on two functions: complexity and connectivity. Plants (vascular) are far more highly connected than animals, and at a cellular level they are as, if not more complex then animal life in cell structure and tissue type. We just don't acknowledge intelligence when its considerations are so distant from our own concerns. I think a revised model where modes of intelligence are attributed starting at an individually cellular model, and build into more complex modes of computational complexity would really clear a lot of this up, and get different modes of intelligence into a singular framework.
- sumitviii 7y ago>>We have some vague models of it ("the brain is like a computer and the mind is like a program running on it") I agree with the thrust of your argument, but this itself is wrong. When the Europeans were figuring out complex gears and stuff, they thought that the brain is like a complex cog machine (that was when they thought that mind was different, I think). When fluid dynamics became the craze, they started comparing, brain to a machine that manipulates some fluid. The truth is that we don't even have a model for it.