The research for my new scifi book (working title is ‘Joe’) is taking me deep into the realm of science, to depths where the scientific imagination begins to bleed into scientific reality.
People are working on some fantastic things that promise to amaze us. Asimov’s robots may be a part of our lives before New Year’s Day, 2029
In 2002 Mitchell Kapor and Ray Kurzweil made a $20K bet on the creation of an AI which could pass the Turing test, with Kapor, the computer engineer, betting against the machine and Kurzweil, the futurist, supporting the proposition. The formal statement of the bet is — “By 2029 no computer – or machine intelligence – will have passed the Turing test.”
The test is administered annually by the Cambridge Center for Behavioral Studies and a Grand Prize of $100K and a gold medal is underwritten by Hugh Loebner. So far no one has won the Grand Prize, though a few have secured awards for near misses. How near? Go have a chat with Mitsuku and see for yourself.
I think she was the 2013 winner, i.e. the chatbot that came closest to passing the test. Like I wrote earlier, there’s not been a Grand Prize winner to date.
Kurzweil argues that the soul question is not relevant. If the AI can perform as a human, that’s a pass. It doesn’t have to actually be human, have a soul, or anything like that. Consciousness not required.
He imagines a reverse engineered human brain built on nano-technology capable of quantum processing.
And that gets to the core of the problem — software.
Kapor argues that no algorithm can ever produce creativity. Kurzweil agrees and proposes a design that mimics the human brain and produces creativity out of sheer complexity.
The argument between Kapor and Kurzweil has been published here.