odysseus2000 wrote:XFool wrote:Again - a common category error - people who think AI means 'Artificial (human) Intelligence'. It doesn't. It means human designed 'Artificial Intelligence'. Not the same thing, apart from some (not very intelligent?) humans who think it is...
The Google engineer saw what he wanted to see - but it wasn't really there. He imagined it was. You're not going to tell me humans don't imagine things or believe in things that are not real, are you? Because if you are, Have I got news for you!
It's all a matter of presentation. Like all the best magic tricks.
I am not sure what you are arguing here.
Intelligence is some innate property of a system that allows the system to solve problems. E.g. a worm, as Darwin showed, as the intelligence to orientate a leaf to the easiest direction for pulling into its hole.
There are examples of animals using tools to achieve various objectives which are clear signs of intelligence.
That's fair enough. But that is a long, long way from the notion that a mechanism is self aware and has insight and understanding - as in LaMDA's alleged "joke".
And I've never heard a worm tell a good joke.
odysseus2000 wrote:There is potentially a debate between intelligence and emotion and whether it is possible to have intelligence without emotion, or whether intelligence without emotion is dangerous and various other combinations.
However, if we stick to intelligence it is possible to do quantitative measurements and to determine if a system has intelligence and although one might layers this at different levels as in e.g. IQ, there is nothing that indicates a difference between a machine or a human in terms of a particular layer of intelligence ability. We now have machines that can play Chess and Go far better than humans with neural net Go and Chess machines able to easily simultaneously defeat multiple human Grand Masters.
Can one argue that these machines in these tests do not have super human intelligence? Or can one say that these machines don't really win and its all a matter of presentation?
This is now taking us into many areas, some very deep and complex and well above my pay grade (and I would suggest above
everyone's pay grade) and some of which are, IMO, risible.
Are these machines "intelligent" ? I think there are two answers: YES and NO. It depends on what is meant by the question. What is being asked?
Obviously many of these systems are now much better than humans at 'playing' these games, I could equally claim that no machine ever really 'plays' Chess, Go or any game. Or, for that matter, can even do simple arithmetic, such as add two simple integer numbers together.
Again, I would suggest tha the word "intelligent" in this context confuses people - to my mind even some clever people - 'artificial intelligence' is
artificial intelligence, not human intelligence brought about artificially. In other words, I don't think more and greater artificial intelligence will, by itself, suddenly become a kind of 'human intelligence' - a being.
This seems to me often less about whether machines can 'think' etc. and more about how humans think (or don't) about things.
1. A simulation of a thing is not the thing itself.
2. Is there a difference between living things and machines? If yes, what is it? (Apart from the obvious)
All this reminds me of several seemingly trivial human misunderstandings(?):
1. People who objected to the the slogan "
Black Lives Matter" by asking: "
Don't white lives matter?".
2. People who objected to the phrase "
Killed by friendly fire" because "
There's nothing friendly about being shot".
3. People who thought scientist carrying out tests on the 'intelligence' of free wild animals in the wild was "cruelty to animals".
Finally, what does certainly remain true, IMO, is: "
There's nowt so queer as folk!"