I was honored to be invited to speak at the Junto, and I'm sorry if my involvement had anything to do with ending that talk series. While some in the audience voiced complaints about the moderation, I was happy to defer to Victor and let him run his talk event however he liked.

I'm pretty sure I didn't give a time estimate of 40 years; I said that the previous history of growth suggested that there might be another transition to a faster growth rate in roughly a century or so. Yes it would be nice to have more data, but we'd be fools not to look at the data we have.

The forecasts in my book The Age of EM: Work, Life and Love When Robots Rule the Earth don't really depend much on when that transition would happen, they mainly depend on our society not changing too much between now and then. My book is also unusually clear, compared to other futurists book, to explicitly express my uncertainty, as overall probability estimates early in the book, and as qualifiers to particular forecasts throughout.

I certainly wouldn't claim that AI would be substantially more able to predict its future than can humans today. They might be able to do this if the world was held constant while their intelligence increased, but smarter AI will make for a more complex and harder to predict world.





Speak your mind

1 Comment so far

  1. Robin Hanson on May 7, 2017 10:45 am

    The text above was from an email of mine responding to text in an email from Victor. I worry that the text above will make less sense to those who haven’t read the text to which it was a response.


Resources & Links