0

At the beginning I have to state that I am a novice and an amateur in philosophy. However I am trying to expand my knowledge and that's why I decided to join this community.

Recently I've came across this paper (by David Chalmers) regarding singularity. I was struck by how his thoughts are similar to mine, but of course better structured and deeply analysed. On the other hand I am surprised that there is no definition of the Intelligence itself (nor the Artificial Intelligence). And yet he freely build statements like

(...) AI that is created by an AIn and is significantly more intelligent than its creator.

How do I know what he means by more intelligent in this context? On intuitive level I am aware what he is trying to say, but I wonder if that is enough to state premises.

That's why I am wondering if I miss something. Is there some well-known philosophical definition of Intelligence (or AI) or we're speaking in broad and vague terms?

Improve this question
1
  • I do not claim that my definition is widely accepted - it is not. But I do claim that I have a better understanding of intelligence than Chalmers. Here is my definition: intelligence is the ability to recognize, process, and cause differences. Note that I do not use tasks, problem-solving, learning, planning, etc. Those are too high-level stuff and will only clutter your thinking. Read Bateson 1972 or Deleuze 1994 if you think that differences do not deserve the attention of philosophers. Details on how differences explain all the cognitive abilities are here - alexandernaumenko.substa
    – user2204238
    Sep 8 at 8:55

1 Answer 1

Reset to default
2

What is generally meant by intelligence would be something along the lines of how humans process and respond to data (be they of sense or of something else) by arriving at novel conclusions. There are a number of things Chalmers might mean by "intelligence" here, but looking at the paper, he's most likely concerned with the ability to solve problems in uniquely useful ways. What's at stake in that paper is the ability of computers to design increasingly more intelligent versions of themselves, to improve presumably both the hardware and the software to enhance the ability to solve problems in ways which would not have occurred to the previous generation of computer and to do so more quickly.

"What is intelligence?" is, beyond that, a tricky question to answer. "When can we say that computers have become sentient?" is almost worse - don't be surprised if you don't see a definition beyond the Turing test by which anyone is willing to stand firm. Most of us (if we're being honest) will say that we're just not sure or that we don't know (though often with the codicil "yet").

Improve this answer
2
  • "with the ability to solve problems in uniquely useful ways". Well, if that's what he mean then it is "just" the "ability to process data" to me. So I suppose that his fears related to singularity are premature. I can easily imagine computer designing more sophisticated computer. In fact that's what we're already able to do in the terms of algorithms. We can create an algorithm that will create an algorithm for the given problem. However if that process is not spontaneous, but supervised there are no risks. Thank you for the clarification I thought I was missing something fundamental here.
    – freefall
    Mar 25, 2016 at 8:25
  • Upvoted for mention of the Turing test which is really all we have at present to check intelligence.
    – Kraang Prime
    Mar 25, 2016 at 9:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .