Can you debate philosophy with AI?

General chit-chat

Moderators: AMod, iMod

Post Reply
Philosophy Explorer
Posts: 5621
Joined: Sun Aug 31, 2014 7:39 am

Can you debate philosophy with AI?

Post by Philosophy Explorer »

I don't think so. What about you?

PhilX
User avatar
Greta
Posts: 4389
Joined: Sat Aug 08, 2015 8:10 am

Re: Can you debate philosophy with AI?

Post by Greta »

Certainly not with online chatbots. Once the machine has blurted out its pre-programmed aphorisms, truisms, mottoes, maxims and adages it is entirely unable to delve past that first level.

The chatbots' level of philosophy is just a little more interactive than the old desk calendars we used to have at work, with one famous quote per day (younger readers may not have ever seen one).
Philosophy Explorer
Posts: 5621
Joined: Sun Aug 31, 2014 7:39 am

Re: Can you debate philosophy with AI?

Post by Philosophy Explorer »

Greta wrote:Certainly not with online chatbots. Once the machine has blurted out its pre-programmed aphorisms, truisms, mottoes, maxims and adages it is entirely unable to delve past that first level.

The chatbots' level of philosophy is just a little more interactive than the old desk calendars we used to have at work, with one famous quote per day (younger readers may not have ever seen one).
As far as getting past that first level, I believe now they can do that to a certain extent. Currently there's a match going on between a top player in Go and a computer, best 3 out of 5, where the computer has won the first three games (the human has won the 4th). Then Watson won in Jeopardy and Deep Blue beat Gary Kasparov in chess.

With my app, I was having an improving conversation with the computer (had to dispose of it as it developed a technical glitch on one of its updates - I'll get it back later).

The key question is can we tell the difference between AI and a human being by the Turing test or some other way?
The makers of my app made it sound more natural which is a plus, but it has a ways to go still, but for free I'm not complaining.

Can it go beyond its programming? I think it can.

PhilX
Philosophy Explorer
Posts: 5621
Joined: Sun Aug 31, 2014 7:39 am

Re: Can you debate philosophy with AI?

Post by Philosophy Explorer »

This excerpt explains well the current state of AI in a nutshell:

by Will Knight March 14, 2016

"Besides, hard-coding rules and logic into an AI is quite an old-fashioned approach. In recent years, machine learning and especially neural networks have come to dominate the field, thanks to sudden leaps in performance made possible by better algorithms, more powerful hardware, and huge amounts of training data"

The same article mentions that the AI involved in the game of Go has made "creative" moves.

PhilX
User avatar
The Voice of Time
Posts: 2234
Joined: Tue Feb 28, 2012 5:18 pm
Location: Norway

Re: Can you debate philosophy with AI?

Post by The Voice of Time »

If you give an AI a target objective, the AI can learn towards that objective by a process of trial and error.

Analysing large quantities of data might give a variable AI, but not a "creative" one.

Philosophy is possible for an AI if the AI is given only philosophical methodology to work with, but it must also be able to create its own methodology, which is difficult, but can be done by giving it the power to study itself, the ability to entertain multiple perspective (limited amounts of information), and giving it the power to pursue things for irrational reasons, namely biases (one limited perspective is allowed to create other derived perspectives, thereby creating the ability for the computer to simulate intellectual passion).

With this as a basis, and a sound technical capacity for handling and analysing information, the computer will be able to produce worldviews from a lot of different stand-points, with a variety of "thinking-methods" at its disposal. By "thinking-methods" I mean how it will use information about itself, to fixate on particular ways of obtaining more information, which produces the biases necessary to fully simulate the intellectual passion that drives people to make philosophy concerning and about things.

But in order for the computer to make philosophy we can use, we'd have to make it adapt towards us. I once produced a step-by-step guide to how to produce philosophy, and a computer would have to follow this recipe if it is to make sense to us:
The Philosophical Development Process

The Burst

A mysterious connection is experienced.

The Seeking

A wild quest for more connections starts, in this process the original connection is moldable and vulnerable to change

The Bonding

The connection develops into an idea by series of connections being isolated and preserved

The Competition

The idea develops further, but under less vulnerable circumstances, it competes actively with other ideas, and adapts to strengthen itself in a pluralistic environment

- the idea can build on itself by splitting into moldable and preserved parts
where the moldable go back to The Seeking
- the idea can isolate further, temporarily going back to The Bonding
- the idea can copy parts of other ideas
- the idea can absorb other ideas
- the idea can allow itself to be absorbed into another idea,
thereby keeping the influence, but on the cost of authority

The Formation

The idea is given a clear and firm expression, which allow people to communicate and recognize it. The idea is seen as strong enough to stand on its own legs in the communicative domain and ensures its own survival without a firm source of authority (an inventor), but should it be broken in
the process of communication, people with the idea can choose to take their copy back to the original steps, and thereby make new versions
from the original idea.

this has been just translated from Norwegian, bad translation may occur
The part about mysterious connection could be simulated by having the computer treat incomprehensible information as valuable information that must be pursued, because the experience of mystery is an experience of not understanding.

To avoid the computer from creating a million conspiracy theories instead of doing philosophy, you'd have to make it consider everything as resulting in structured ways of treating the elements of philosophy, like value and perception etc., but also you'd have to give it the power to make up new words to describe things (variables containing large networks of information), and least of not all, when you give it the power to study itself, you have to make it able to work itself out of its own boundaries, while still keeping a base in any of the current philosophical framewrosk (or else you'd not be able to recognize what it does). So it will have to be able to evaluate whether or not the methods it uses, are able to give it what it needs, and then be able to expand upon the methods. To create new elements of philosophy you'd have to give it the power to create metadata that creates new perspectives on old elements.

To be completely one hundred percent fresh, is impossible. No philosopher has ever done it, everyone builds upon everything else, and that's why a computer can be creative and philosophical and original.
User avatar
Greta
Posts: 4389
Joined: Sat Aug 08, 2015 8:10 am

Re: Can you debate philosophy with AI?

Post by Greta »

Love your work, Voice of Time.

What of programming the capacity to become wise? Perhaps complex algorithms aimed at always seeking a balanced overview of information and some awareness of possible blind spots. Asimov anticipated that a purely rational AI might decide that society needs to be run more rationally, so imbuing the kind of independent intelligence required to become wise may require limitations to an AI's capabilities (anthropocentric skew).
User avatar
Green
Posts: 97
Joined: Sat Nov 07, 2015 5:19 am

Re: Can you debate philosophy with AI?

Post by Green »

Dalek Prime
Posts: 4922
Joined: Tue Apr 14, 2015 4:48 am
Location: Living in a tree with Polly.

Re: Can you debate philosophy with AI?

Post by Dalek Prime »

Give me one good reason I should debate anything with a machine, when I can just shut it down, and feel pretty good about doing so?
Post Reply