Hacker News Reader

Comment by CuriouslyC | Hacker News Reader

Artificial general intelligence won't be invented, it will emerge, just general intelligence did in the wild. AGI is just going to be a hierarchical arrangement of specialized tools.

The first step is highly specialized AI tools for very specific problem domains.

The second step is AI tools that use other AI tools as components but address slightly larger problem domains.

The third and successive steps are recursions of the second step.

Additionally, we won't be able to tell right away when we've crossed the threshold. We can't even say for sure where "intelligence" stops in animals. We used to think we were the line, but now the bar has been pushed down to include primates, cetaceans, a number of birds, possibly some members of the bear family, etc. The reality is that it is a gradient and there is no clear line.

View on HN

> Artificial general intelligence won't be invented, it will emerge, just general intelligence did in the wild.

I think so as well, but for a specific reason: That people are not as "AGI" as we think, and that our general intelligence is the result of immense neuron-computing resources committed to satisfying a few simple drives. In other words, narrow AI with a lot of resources. Not to mention many specialized subcomponents.

And to the degree that consciousness is a selected trait, that means it has a purpose. I think the work on attention/focus in neural networks hints at that purpose.

Why do you think consciousness is a selected trait?

Well I don't know what consciousness is, so I couldn't answer that. But I think the opposite stance of consciousness being some kind of majestic oracle is silly. There is a reason why we perceive consciousness, even if it is just as a side effect of something else or a random fluke that is useless or an illusion for coordinating our limbs or whatever.

These types of discussions do get pretty wonky. When I say consciousness is selected for, I mean more whatever practical apparatus (if any) it is emergent from, not the perception of consciousness itself.

Do we perceive consciousness? Or is consciousness perception itself?

It sounds to me like you're referring to cognition rather than consciousness.

It's interesting how people come out of the woodwork with their personal theories on AGI. Do you/ we even really know how general intelligence works, or even how it emerged i.e. incrementally, or in a dramatic mutation more recently? Last I checked there wasn't a scientific consensus on either topic. To then come out headstrong and say "AGI will be like X" always makes these AGI conversations a tad farcical.

Intelligence clearly isn't a one-off/recent thing, since we observe remarkably intelligent behavior from cephalopods, which are vastly distant in the tree of life and not recent from an evolutionary perspective. We also know intelligence is also clearly not a binary attribute from many animal and human studies.

The fact that we don't know how higher-order intelligence works in general is exactly why it will be emergent rather than designed.

You shouldn't worry so much about consensus, but instead use your senses and your brain to make up your own mind. That is the approach that gave us the enlightenment.

Ok but I am talking about general human intelligence when I say AGI and 'recently emerged', not mollusks. You could easily argue we will find out how higher-order intelligence emerged one day, as some researchers already have models for that if you've read a college anthropology textbook; may not be right, but it's not out of the realm of possibility that it emerged recently due to new structure(s) ('design') coming about in a relatively short time period, so making the claim "Since we don't know how it happened, therefore it must be like X" is flawed.

You know what French Enlightenment thinkers were also against? Making headstrong claims (as an authority) without empirical evidence on your side :)

I think "what humans are" is not a useful definition of intelligence. Not only is it not useful, but it's likely to lead us down a blind alley in intelligent systems research. Kind of like if we assumed when trying to build a flying machine that the only way it could work is if it flapped its wings.

Human intelligence definitely did not arise "in one day" and any anthropologist positing such a theory hasn't looked at the last 30 years of research in statistical genetics. Intelligence is an incredibly complex trait resulting from the interaction of many, many genes.

Surely it isn't too much trouble to mentally insert "My intuition is that" before the GP's interesting comment. When dealing with a hard problem, whose solution is many steps removed, we routinely use educated intuition to guide us. It seems to me that insisting on empiricism in this space, at this time, would practically mean halting most thought on the topic.

This. This i believe is the how its going to turn out. Specialized AIs cooperating and creating higher level abstractions. It may not be AGI at first but Practical AI that could be used for real world applications where human intelligence is usually needed.

Reply on news.ycombinator.com