Panu Horsmalahti wrote:
In my opinion the definition of intelligence is a machine that has an input and an output. It covers all the intelligent systems you can imagine, but also things like thermostats. I think it is too human centric to think that a thermostat is not intelligent because it doesn't have legs and it can't do much. It is very stupid (if you think it is possible to rate intelligence), but still, it can be viewed as an intelligent system. Intelligence really means many, many things and I'm not really that interested in a strict mathematical definition. It reminds me of all the philosophers who have debated for thousands of years on the definition of love, art and such. In the end, no definition is more correct than any other. You can build an AGI without really knowing what intelligence is, just like you can paint beautiful art without knowing some strict definition of art.

Ah, but you can do one of two things here.

You can say "defining intelligence is actually so tricky that nobody can do it".

That's good: not only do I agree with this, but I can come with an argument as to *why* it should not be possible.

But if you then go on to say: "intelligence is a machine that has an input and an output" don't I have the right to say "Huh? You just said
it could not be done!"

People are looking for definitions of intelligence because IF they could nail down a strict (quasi-mathematical, rigorous) definition, they could then go on to say that the next step is to do some work that clearly and unambiguously IMPLEMENTS that definition. It gives a kind of stamp of authority on their AI efforts.

For example, the statement that "An AI is an agent that tries to satisfy a set of goals" (or some such wording) seems to make the AI endeavor look like it starts from a clear conceptual beginning that DOES NOT REFER to any attempt to copy human intelligence. This is why the agent-talk definition is required. It is a political tool, of you will.

But it is nonsense.

Richard Loosemore.

-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?member_id=231415&user_secret=fabd7936

Reply via email to