According to Tristan Harris, facing the “AI dilemma” of both its power and its danger, “we need to move at the speed of getting it right”.
We’ve featured the Centre for Humane Technology. and its founder Tristan Harris, for a number of years now (see archive). Notable for the documentary The Social Dilemma, they are hardly luddites vis-a-vis computation and digitality - indeed Harris’s co-founder, Aza Raskin, is using AI as a kind of rosetta stone to begin communication with animals (Earth Species Project).
In a presentation last month (March), Harris and Raskin turned their attention from the toxicity of social media, to the potential dangers of an untrammelled AI, in its current Large Language Model forms. “The AI dilemma” is embedded in full above - and there’s a podcast/transcript with the same title, though it’s a private and separate dialogue between the two.
At AG, we are generally open-minded about the possibilities for human and communal empowerment that come from advanced technology. But this is a somewhat cautionary call, comparing the rise of superintelligent AI to the advent of nuclear weapons, in terms of the challenges of design, control and safety. Worth hearing.
And see this week’s editorial for further musing on AI’s promise and peril.