Charbel-Raphaël comments on Constructability: Plainly-coded AGIs may be feasible in the near future

[Charbel-Raphaël]

Here is the youtube video from the Guaranteed Safe AI Seminars: