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

A conversation with Sussman on AI and asynchronous programming

dustycloud.org

Post reply on HN
Print view View original

11–20 of 35 posts

Re: A conversation with Sussman on AI and asynchronous programming

#11

Post by davexunit »

kruhft wrote: post #10

Whenever I'm out of ideas I always end up going back to the Propagator/Logic Simulator chapters of SICP and rewriting them hoping to glean a bit more about the process and usability of such an abstraction. I usually end up writing a Propagating logic simulator (combining the chapters) and marveling at the working forward/backward ripple carry adder. This time I've derived all the way down to switches which infer unkn…

I took that chapter and went in a different direction, making the network a DAG and building a functional reactive programming environment out of it.

Now, to go back and the propagator network and read the Scheme-Propagators source.

Re: A conversation with Sussman on AI and asynchronous programming

#12

Post by nalaginrut »

When you put all things into a blackbox and pray for the promising result, the whole methodology goes to a dead end because it's not revealing the truth but voodoo magic to self-cheating. Nowadays, people just put big data into a blackbox to make "AI". We need the AI system to tell us why, not how. Or it's not AI to help you, you made yourself a Lord to give you order. And you never have chance to ask it why.

Symbolic AI could be a chance, because it keeps necessary meta info to reveal reasonable things. But how to optimize it properly will be a problem.

Re: A conversation with Sussman on AI and asynchronous programming

#13

Post by themgt »

"How am I talking to you right now?" Sussman asked. Sussman seemed to be talking about the shared symbolic values being held in the conversation, and at this point I started to understand. "Sure, when you're running the program, that whole thing is a black box. So is your brain. But you can explain to me the reasoning of why you did something. At that point, being able to inspect the symbolic reasoning of the system is all you have." And, Sussman explained, the propagator model carries its symbolic reasoning along with it.

I dunno. Are we talking about true "hard AI"? Because really, how much can Sussman or anyone introspect their own thought processes? We can tell ourselves stories about motivation and symbol manipulation, but those stories themselves are just more symbol manipulation. I'm not sure a sentient symbol-manipulator ever really "gets to the bottom" of true introspection of their own thought processes, including counterfactual thoughts they might have had.

This seems like an overly mechanistic approach, although perhaps workable in "soft AI"

Re: A conversation with Sussman on AI and asynchronous programming

#14

Post by xzephyr »

kcorbitt wrote: post #8

There definitely exist "accountable" AI models. Things like decision trees and various types of regressions. The thing is though... any sufficiently advanced AI is going to be unaccountable pretty much by definition. It's like, calculus is an extremely useful tool for predicting the temperature of a cooling object over time, but good luck explaining to a 3-year-old how to perform the necessary maths. The fact that th…

Seconded. Another extreme example would be human brains, which I don't think we understand enough for it to be "accountable" in any mathematical bounds, yet we trust them to make complex decisions. Statistical characterization of the behavior of the AI system is a better objective than the ones based on inherently biased symbolic systems. Just because it is the way how humans communicate it, doesn't mean it's the best way to do it.

Re: A conversation with Sussman on AI and asynchronous programming

#15

Post by deckard1 »

themgt wrote: post #13

"How am I talking to you right now?" Sussman asked. Sussman seemed to be talking about the shared symbolic values being held in the conversation, and at this point I started to understand. "Sure, when you're running the program, that whole thing is a black box. So is your brain. But you can explain to me the reasoning of why you did something. At that point, being able to inspect the symbolic reasoning of the system…

My thoughts as well.

This is why I feel some people will never "let" AI happen. They want 100% certainty. Neural nets scare them. Despite the fact that real, human intelligence carries no certainty. You either have intelligence, and creativity, and uncertainty, and accidents. Or you have computers and rigid logic.

The interesting thing there is, when a space rocket malfunctions, when a car wrecks, or the stock market goes crazy... and it's a computer involved, it's always human error at the end of it. And, thus, intelligence that caused the problem. Computers mindless do exactly the thing they were given to do.

Re: A conversation with Sussman on AI and asynchronous programming

#16

Post by PuercoPop »

themgt wrote: post #13

"How am I talking to you right now?" Sussman asked. Sussman seemed to be talking about the shared symbolic values being held in the conversation, and at this point I started to understand. "Sure, when you're running the program, that whole thing is a black box. So is your brain. But you can explain to me the reasoning of why you did something. At that point, being able to inspect the symbolic reasoning of the system…

It seems more inline with Justification based/Truth maintenance systems than hard AI.

Re: A conversation with Sussman on AI and asynchronous programming

#17

Post by rurban »

paroneayea wrote: post #5
rurban wrote: post #2

This accountable argument is stone-old. You'll find it in an intro AI book. Of course only when implemented in lisp, so it needs be an older book. It's the typical argument against neural nets, because they cannot explain their chain of reasoning, and so you are not able to train it to right way, or do not train into the wrong direction. when something goes wrong you got a problem. Old AI had the same problem, that's…

Maybe true, but "stone-old" ideas don't mean "bad ideas". Neural networks were "stone old" until all this big data stuff went crazy and now suddenly they're on the frontpage of hacker news all the time again and people think they're the New Hotness (TM). Similar with many long-forgotten functional programming techniques which are being talked about as new stuff, much of it is refinements of old ideas, but finally hit…

When I mean "stone old" I mean of corse better than todays ideas. This should be clear from the context. Not everything which is fast is also good.

You can also come up with some kind Greenspun's tenth rule applied to neural net's:

"Any sufficiently complicated modern AI program contains an ad hoc, informally-specified, bug-ridden, slow implementation of half of Common Lisp."

In this case not even that.

Re: A conversation with Sussman on AI and asynchronous programming

#18

Post by marcosdumay »

rspeer wrote: post #7
paroneayea wrote: post #5

Earlier quoted context omitted.

Maybe true, but "stone-old" ideas don't mean "bad ideas". Neural networks were "stone old" until all this big data stuff went crazy and now suddenly they're on the frontpage of hacker news all the time again and people think they're the New Hotness (TM). Similar with many long-forgotten functional programming techniques which are being talked about as new stuff, much of it is refinements of old ideas, but finally hit…

My dream would be if Good Old-Fashioned AI (the symbolic, explainable kind Sussman is interested in) were to have the sudden redemption that neural nets had. It was easier for neural nets, though, because they were close to a previously successful AI mechanism (machine learning with logistic regression), it's just that we had spent decades talking about them with different words for no good reason. There's a much lar…

Pattern matching has unbeatable performance benefits, at least for our current computers.

If you go deep in numerical calculus, you'll see that our computers are must better fit to work with continuous smooth functions than they are for discrete noise-like data. So, all the power goes to the people that turn their knowledge into a smaller set of interpolated curves. (And yes, I think that's very counter intuitive.)

Anyway, I'm not convinced this is a fundamental property of CS. It sounds at least possible that different architectures could have different constraints, and make symbolic AI viable. But it those would need to be very different architectures, not based on the Von Neuman or Turing's ideas at all.

Re: A conversation with Sussman on AI and asynchronous programming

#20

Post by panic »

themgt wrote: post #13

"How am I talking to you right now?" Sussman asked. Sussman seemed to be talking about the shared symbolic values being held in the conversation, and at this point I started to understand. "Sure, when you're running the program, that whole thing is a black box. So is your brain. But you can explain to me the reasoning of why you did something. At that point, being able to inspect the symbolic reasoning of the system…

You can usually come up with an explanation for why you did something. You don't have to "get to the bottom" of your own thought processes to do this: you just need to be able to reconstruct the symbolic manipulation part. This seems like a good thing for an AI to be able to do -- especially a truly "hard" AI that you'd trust to run things at a high level.
Post reply on HN