Bill Gates, Musk, Hawkings and others have all stated their concern regarding the distressing potential dangers of AI more than once, on we go pell-mell towards self aware/self governing machines.
We can’t even get security updates right without causing severe problems, but somehow think “We can do this. We can win!”.
Just a minor thought…a program, any program (including heuristic ones) are limited by their coding and how, via this coding, they ‘learn’. The same is true about biological systems. Their form, their being carbon based, their being subject to the laws of thermodynamics and sensitivities to environment, and other biological entities all determine and limit how they learn.
Another minor thought, “If something can go wrong, it will.” Just ask God.
Now comes this report by Selmer Bringsfjord (RPI, New York) (inThe New Scientist) regarding a test he ran using the classic “Wise-men Puzzle” on three robots, two of which he silenced, and one he didn’t. All three had auditory sensors.
“In a robotics lab on the eastern bank of the Hudson River, New York, three small humanoid robots have a conundrum to solve.
They are told that two of them have been given a “dumbing pill” that stops them talking. In reality the push of a button has silenced them, but none of them knows which one is still able to speak. That’s what they have to work out.
Unable to solve the problem, the robots all attempt to say “I don’t know”. But only one of them makes any noise. Hearing its own robotic voice, it understands that it cannot have been silenced. “Sorry, I know now! I was able to prove that I was not given a dumbing pill,” it says. It then writes a formal mathematical proof and saves it to its memory to prove it has understood.” – New Scientist
Granted, this isn’t “full consciousness”, but this is conscious thought and shows a conception of ‘self’, or “the first-hand experience of conscious thought’.
There are those who are correct in saying that there’s a big difference between saying, “It’s sunrise.” and being able to enjoy the esthetic experience of knowing who you are and being part of that sunrise and . Perhaps central to the experience is knowing one is mortal and what that sunrise signifies in terms of mortality and the passage of time which generates compassion for others subject to that passage of time and the knowledge that each is at a different point in that passage.
Perhaps what I fear most therefore, is a machine which has no compassion and its actions for self preservation without that essential quality, even if through inaction because it simply has no perception that it is doing wrong since ‘right’ and ‘wrong’ are alien to it.
After all, even though very imperfect, we do have a system of checks and balances, ideas of morality, etc. which function (to some degree) to limit us.
If you don’t believe the craziness of all this, if you don’t believe this is real, read about how ‘killer robots’ was to be discussed at the the U.N. Convention on Certain Conventional Weapons. You can look up the meeting (11/2014) search. You read more here.
Source:
https://www.newscientist.com/article/mg22730302-700-robot-homes-in-on-consciousness-by-passing-self-awareness-test/
http://www.computerworld.com/article/2970737/emerging-technology/are-we-safe-from-self-aware-robots.html
http://www.stopkillerrobots.org/2015/03/ccwexperts2015/
http://www.computerworld.com/article/2489408/computer-hardware/evan-schuman--killer-robots--what-could-go-wrong--oh--yeah----.html