Cito di Pense wrote:_exit wrote:Did you know that
AGI work seeks to replicate something in humans called "
general intelligence", regardless of how the word consciousness may be varyingly defined (or rather, regardless of how the word consciousness may be varyingly understood)?
Well, the wikipedia article on AGI does not start out well:
Artificial general intelligence (AGI) is the intelligence of a machine that could successfully perform any intellectual task that a human being can.
That use of
any doesn't really describe the task of anyone doing research on AI, so really, the search for AGI is related to a set of tasks that somebody thinks represents a useful kind of AI. Useful to whom? Why, humans of course. With only humans to define the target 'intelligence', don't you think the language of 'general intelligence' is a bit grandiose? Some example tasks are listed a little later in the article, but such an article is only really pertinent to a non-expert who might want to pretend to understand what actual AI researchers are up to.
In sum, you're not after anything technical, here. In fact, the thread title you chose (related to the Bennett article you linked) is bordering on absurd. The problem of
purpose is even worse than the problem of
intelligence. People who want to get something done stick to words like
function, which will get you back to focusing on the kinds of tasks you want the machine to perform. Who sets those priorities? Your thread is in the wrong subforum, because you're just asking philosophical questions. There's nothing wrong with that, but putting this kind of discussion into a technology subforum is also bordering on the absurd. There are very few folks contributing to threads like this who know very much about technology, anyway.
I'm just not seeing how your response above offers anything reasonable to the discussion; it doesn't change the fact that some of the
planet's smartest people are working to replicate the human mind in inorganic form, and it doesn't change that AGI may need be humanity's final invention!Also, as far as we can see, humans are things with quite profound general intelligence (See Wikipedia Neuroscience), so we base AGI's research goal on that outcome, although we still refer to other intelligences on the way to human like outcome.
(i.e. C elegans' intelligence/neuronal system has appropriately already been replicated, in an attempt to help to understand more complicated intelligences.. So artificial intelligence may occur in a gradation of intelligences, until human level is reached)