During the first half of the show, Paul Hynek (son of the famous Project Blue Book scientist) joined guest host Jimmy Church (Twitter) to take a deep dive into artificial intelligence (AI). Hynek defined AI as an intelligent agent that can perceive its environment and continually adapt to maximize the probability of success, while instantaneously conceptualizing different visions of the future to match its actions to the best chance of success. This is what separates AI from regular software. He outlined three kinds of AI: weak or narrow (Siri, Alexa, driverless cars), strong or general (can pass Turing Test), and super intelligence (greater than human capabilities).
Hynek commented on whether it matters if we are characters in an advanced extraterrestrial intelligence's simulation. "You probably wouldn't even know... it would almost be like a guardian angel, you know, there's somebody definitely there making the rules," he explained, noting the freewill so many are concerned about not having is largely an illusory concept. He suggested the singularity resulting in the rise of AI self-awareness will be an ongoing process beginning with adding machine capabilities to our existing consciousness, rather than machines developing consciousness on their own. "That can and, I believe, will produce the singularity," he said.
------------------------------------------------
In the second half of the program, futurist Isaac Arthur continued the discussion on artificial intelligence. "We definitely have stuff to worry about... so long as we're reasonably cautious we can probably get through it and come out ahead," he cautioned. According to Arthur, there are enormous obstacles to overcome in order to go from AI of the kind we have today to super intelligence. "We've been trying to make humans smarter for centuries, we don't really make much progress," he quipped. Arthur pointed out we have AI systems but they are only about as smart as an ant so he does not worry about rapidly accelerating AI in the near future.
It does not matter if humans are in an advanced extraterrestrial intelligence's simulation as this is the only reality we know and reality can be defined by what one experiences, he explained. "Real is what you make of it," he added. Arthur suggested AI will one day know that it is a simulated intelligence but will not likely know if the reality that it is experiencing is real. He covered power consumption issues of supercomputers that use a megawatt to run roughly the same processing power as a human brain (which only uses 15 watts). Arthur forecasted humans may have significantly advanced AI technology within a thousand years, maybe even something on the level of a Matrioshka Brain. Such an enormous AI system utilizing the full energy output of a star could simulate reality and run a simulation of the universe.