ndanger.organism

It is also important to note that once a computer does achieve a human level of intelligence, it will necessarily soar past it.
– Ray Kurzweil

Ray Kurzweil and Mitch Kapor debate the distant future of the Turing Test. Kurzweil–who I agree with more than Kapor–bases his argument on exponentials. Once self-aware machines exist they will improve rapidly for two reasons: 1) short iterations and 2) directed evolution.

If we assume that a computer that is self aware will take 1 micro-second (10^-6) to reprogram itself, one human generation (say 20 years) is more than 6 x 10^14 machine generations. Put another way, humans have been around for 2 million (2 x 10^6) years, or 10,000 generations. Machines would go through 10,000 generations in less than a second.

And for humans that was 10,000 generations of trial and error! Machines can direct their evolution. It’s true we can direct our technology, but imagine being able to, say, reprogram yourself to be 10 times smarter (or 10 times more beautiful which is what we’d probably all choose instead).

Combine the two and as soon as we develop machine intelligence bam! singularity. The very definition of the singularity is that we can’t predict beyond it.

To hell with that.

The lifeforms of the future–a more apt term than artificial intelligence–will be as different from us as we are to mice. Or perhaps as different from us as we are to current AI’s.

Maybe in 2078 some pundits will be sitting around arguing if humans can pass the ServoTron8000 Test. “No, ” one will argue “although the humans have a demonstrable grasp of intuition, they have never shown anything like full fledged holo-logical thought processes.”

I, for one, welcome our self-reprogramming digital overlords.