LW - From Deep Learning to Constructability: Plainly-coded AGIs may be feasible in the near future by Épiphanie Gédéon | The Nonlinear Library | Podwise