Monday, 16 September 2019

Is Artificial Intelligence alive and conscious; and, if so, how?

I regard the notion of Artificial Intelligence (AI) as manipulative or delusional, built upon many layers of incompetence and dishonesty.

But there is a fair question that could be put to someone like myself who believes that Everything is alive and conscious: reality is made-up of Beings.

Since Everything (humans, mice, trees, mountains, planets) is - in some sense - alive and conscious; then why not computers?

My answer is that computers Must be alive and conscious in some sense; but it all hinges on how.

Some things are Beings, and other things are parts of Beings. Humans are Beings. But most of the things we commonly regard as not-alive and not-conscious are parts-of Beings - things like the cells of a coral, the atoms of a grain of sand... These are probably not Beings in their own right, but participate in larger, more inclusive Beings that do have some degree of bounded, agent consciousness.

So, it is in this sense that computers are alive and conscious - they are part-of Beings; not Beings in their own right. Computers can participate in systems of intelligence, but cannot themselves be intelligent.

And these intelligent systems are Beings (at least, they are Beings by a minimalist definition).

The danger of AI is therefore Not that computers will become Beings, agents, conscious and alive; but that we will begin to treat them as such. This would be intrinsically evil, and lead to greater evil.

The Turing Test is an example of evil propaganda. It argues and implies, rhetorically, that anything that superficially appears to be a Being by reductionist and simplified criteria; really is a Being, and ought to be treated as a Being.

This treating of a part-of-Being asif it were a Being would actually be a human choice - and would be a chosen-lie, a falsehood, Not-Really Real - thus it would clear a path to the condition of virtuality; in which the not-real it lived as if it were real; and the real is ignored or denied.

Which is precisely where we are now.


Bruce said...

William Briggs had a recent post on AI you might enjoy.

a_probst said...

Turing may have posed the test with a bit of professional bravado, and perhaps with a view to tweaking the noses of his employers who had treated him shabbily.

Maybe it should have been called the Turing Conundrum or the Turing Warning. A too-cunning imitation of humanity, repeatedly encountered and dealt with, may not only lead to an over-valuation of the artificial but to a devaluing of humanity much in the same way we grow inured to simulations of murder on television.

Alive and conscious? No. Like many I unthinkingly enjoyed tales of robots in Asimov, Star Wars, Douglas Adams, etc., regarding them as characters. But seeing video of the NASA robot chilled that out of me instantly as far as reality is concerned. When one deals with a humaniform robot (android in science fiction jargon) one deals not with a person but a committee: psychologists, cyberneticists, software designers, even marketing consultants, all of them engineering the thing to engender certain responses from real people.