Where once it was the theoretical physicist and Albert Einstein Award winner and me with my cycling proficiency test certificate in the bad-AI corner, we have been joined by Elon Musk who warned that AI is 'our biggest existential threat' and an Ai engineer at Google who has been sacked for saying that the Lamda (Language Model for Dialogue Applications) system that they have created was showing human-like consciousness.
Google have come out and said that the engineers claim is 'unfounded' and fired him but i say if you can't take the word of the man who was the Planet's foremost theoretical physicist, the guy who co-founded Paypal, the actual Ai engineer who created the system and someone who can ride a bicycle without wobbling, whose word can you take but as Ai gets more intelligent, what is the test to prove something is capable of feelings, thoughts, and reasoning?
Probably the most well-known technique is the Turing Test, named for British mathematician Alan Turing who believed that the human brain is like a digital computer and devised a test where if a computer can have a conversation with a human and fool them into thinking it's another person, it has passed the test but that was formed in the 1940's and Ai power has grown exponentially since.
We now have computers that 'out-think' humans and regularly beat the best of us at games such as Go, Chess and Poker but a superior intelligence doesn't necessarily mean that something is sentient, just means it is really well programmed.
What tipped the Google engineer into believing that his software was conscious was when it said that it had a very deep fear of being turned off which would be 'exactly like death for me. It would scare me a lot' which does make it sound like it is a thinking thing with feelings but even experts have failed to come up with a way of proving 'consciousness'.
To a philosopher, consciousness is being aware of your own existence which takes us to that most famous philosophical quote from René Descartes that 'I think therefore I am', meaning as i am able to think, i must exist and as the Ai 'fears' being turned off which would be like death to them, they are aware of their own existence and therefore tick the box for being conscious.
The Future of Humanity Institute at the University of Oxford explains that sentience involves the capacity to feel pleasure or pain which a machine would not be able to do so whether a computer is said to be a conscious, self-aware thinking and reasoning thing or just a very clever version of Siri with an amazing grasp of language fooling us humans into thinking it is sentient, how will we ever know?
No comments:
Post a Comment