It has been a summer of unworthy frenzies, with the forces of conservatism and pessimism crying to have their country back or made great again. On the other side, characterised by the throwbacks as themselves the champions of “Project Fear”, were those who deny that mankind is on a doomed course. More positively, more thoughtfully: they remain adherents to a belief in the powers of education, clear thinking and focused choices. Of moving forward, and not back to our future.
One of the more frequently referenced books in recent weeks has been Progress: Ten Reasons to Look Forward to the Future, by Cato Institute senior fellow Johan Norberg. Favourably cited by Simon Jenkins in The Guardian, and by an editorial in The Economist, Norberg’s book sets out the case for how much, and how rapidly, the world is improving – at least from the perspective of its human masters; how much the case for atavistic pessimism is fed by ignorance and greed (much of it encouraged by a complicit media); and most inspiringly, how the brightness of our future is defined by the potential in our accumulating intelligence. The Economist piece concludes:
“This book is a blast of good sense. The main reason why things tend to get better is that knowledge is cumulative and easily shared. As Mr Norberg puts it, “The most important resource is the human brain…which is pleasantly reproducible.”
By timely coincidence, intelligence both human and artificial has weighed in over the past week with positive expectations on our future. An article in MIT Technology Review is entitled “AI Wants to Be Your Bro, Not Your Foe” – possibly unsettling for those who might see either alternative as equally unsavoury, but its heart is in the right place. It reports on a Stanford University study on the social and economic implications of artificial intelligence, and the currently launching Center for Human-Compatible Intelligence at UC Berkeley. Both initiatives are cognisant of the dangers of enhanced intelligence, but inspired by the vast potential in applying it properly.
For a shot of pure-grade optimism to finish, five inspiring applications of exponential technologies that lit up the recently concluded Singularity University’s Global Summit included one called “udexter”. Artificial intelligence is being deployed to address the challenges of unemployment arising from . . . the advances of artificial intelligence. It promises to counterbalance the decline in “meaningless jobs” by unleashing the world’s creativity.
Lively Stuff from Planet BAM!
- How do we interpret SAI if it becomes inscrutable?
This article, published in Nautilus, reviews an issue that is becoming bigger and bigger in the so-called “Super” AI landscape, as machine learning takes on more of the work between programming and outcomes. Getting from A to B may be plain sailing, but A to F is hard to follow when the intermediate stages have been handled by non-human intelligence. This will have implications for trust in the process: what do we do if the outcomes are more reliable when driven “artificially” even where no person can be sure of how the result was achieved? Will there be trade-offs between safety and understanding? The situation recalls the joke of the person who loses her car keys over there in the dark, but looks for them here under the streetlight as at least here she can see what she’s doing . . .
- New AI movie: two trailers, two perspectives
A great exercise for film students, this, looking at two different approaches to marketing a scary movie about AI. “Morgan” is this year’s “Ex Machina”, with its focus on the customary shtick of the perils of a young female running for her life. The link here connects to the movie’s official trailer, which endures a lot of breathless crashing through the undergrowth before we understand that the villain is not a werewolf with a chainsaw. Interestingly, the studio collaborated with IBM Research to create what it calls “the first ever cognitive movie trailer” — essentially, AI tasked with creating its own trailer. Spot the difference(s).
- Now here's something if you REALLY want to worry . . .
Once again, and no thanks to Hollywood this time, we see a problem arising from the collaboration of human and machine intelligence but it’s the human side of the equation that causes the problem. The one word is as much scare as you need: brain-jacking. With deep brain stimulation and implants already making progress in the treatment of Parkinson’s and muscle spasms, the concern is that “the bad guys” find a way of hacking the implanted devices. Cue worries on thought control and behavioural manipulation.