Realistically, how far away are we from human level AI?

The Something Awful Forums > Discussion > Debate & Discussion > Realistically, how far away are we from human level AI?

twodot: Aug 7, 2005; You are objectively correct that this person is dumb and has said dumb things

Thug Lessons posted:

You're confused as to what general intelligence is. It's not a philosophical statement about the nature of intelligence, it's just a best-fit line that represents covariance of the results of cognitive tasks, and so far we've found only two tasks that g fails to predict reliably: athletic and 'musical' tasks.

How well does g predict the ability to render H264 video or the ability to store a trillion bytes of information in long term memory? Comparing human tasks to computer tasks just doesn't make any sense. If there's some reason we need more than 7 billion humans (we don't), we already know how to build more.

# ¿ Nov 27, 2016 20:11

Adbot: ADBOT LOVES YOU

# ¿ May 14, 2024 02:13

twodot: Aug 7, 2005; You are objectively correct that this person is dumb and has said dumb things

Elias_Maluco posted:

Im on the "we are not even close" field. As many others have said, we barely know how our own intelligence works

We don't need to know how a bee flies to build to planes.

quote:

The little we actually knows, it seems like our intelligence doenst works at all like a computer. All a computer can do, the way its built right now, no matter how advanced and powerful it is, is basically mathematics over data. Out brain is weak in performing mathematical calculus, the machine has been doing it a lot better since the beginning. But our brains do a lot more than mathematics

This implies there something we can do that can't be expressed in mathematics, which I've seen no evidence of. But more than that, you've chosen an abstraction layer for computers that just doesn't apply to humans, and it's causing you to come to weird computers. What computers do is shunt electrical signals around in fixed patterns, for the convenience of humans we've organized those shunts into logic gates, and those logic gates into adders, registers, and such, and those components into a CPU and memory. The result is a thing that takes instructions like "add this to that" or "go run that instruction" which looks sort of like mathematics, but there's absolutely no reason a computer needs to be organized like that (other than that humans would have a hard time understanding how to use it). Neural networks are essentially an expression of that fact, but implemented in software.

The fact that it's even possible to run dynamic shunting software on static shunting hardware is because, as near as we can tell, computers are general purpose problem solvers, they're Turing complete, and we've haven't encountered a solvable problem that isn't decidable by a Turing machine (even if a particular machine is inefficient compared to other machines). Humans weren't designed so it's harder to crack open a skull and say "Ah ha! Here's the adder" (though people certainly try), so we can't apply your computer abstraction level to humans, but they look pretty similar at the electrical/chemical shunting machine level.

twodot fucked around with this message at 19:12 on Nov 29, 2016

# ¿ Nov 29, 2016 19:09

twodot: Aug 7, 2005; You are objectively correct that this person is dumb and has said dumb things

Condiv posted:

uh, yeah you can dude. ever followed a recipe and made some food? congrats, you were just programmed in a natural language

When I follow a recipe there's large probabilities I will gently caress it up. And not just gently caress it up like "Oh this could use some salt", but I could drink too much, pass out, and burn down the neighborhood. If we're expecting that sort of accuracy from computers, I'd say we've already got natural language programming.

# ¿ Nov 29, 2016 20:03

twodot: Aug 7, 2005; You are objectively correct that this person is dumb and has said dumb things

Condiv posted:

it's not about accuracy. being able to process instructions that are conveyed like normal human language is the point of true natural language programming

How the gently caress do you plan on measuring on whether instructions that are conveyed like normal human language were processed other than accuracy? If I give you a computer that takes your input, gives it to Cortana, and then either does what Cortana says or burns your house down, how do you know the computer didn't process your natural language instructions and just got drunk instead of doing what you asked for?

# ¿ Nov 29, 2016 20:13

twodot: Aug 7, 2005; You are objectively correct that this person is dumb and has said dumb things

Subjunctive posted:

Yes, I know. They were two examples.

Without any sort of knowledge about "memnets", it looks pretty weak to offer up two examples, and totally fail to defend one example, and just say "but my other example was the good one". It reads like a lazy gish gallop.

# ¿ Dec 23, 2016 02:38

Adbot: ADBOT LOVES YOU

# ¿ May 14, 2024 02:13

twodot: Aug 7, 2005; You are objectively correct that this person is dumb and has said dumb things

Subjunctive posted:

I was asking a question in earnest -- did he not consider them to be meaningful advances. I wasn't asserting anything about what he should believe. Sometimes a question is just a question, not a Socratic feint.

What do you care about random Internet person's opinions other than to make a point? Either they're aware of the stuff you mentioned, and they by definition don't consider them meaningful advances in mathematics, since they've already stated their position, or they're not aware, and you're trying to score points by having better jargon than them. (seriously "memnet" turned up nothing obviously related) The way to phrase that question earnestly is "Why don't you consider memnets to be a meaningful advance in mathematics?"

# ¿ Dec 23, 2016 02:53

The Something Awful Forums > Discussion > Debate & Discussion > Realistically, how far away are we from human level AI?