Comment by ysavir

Seems the opposite way round to me. We couldn't conclusively say that AGI is possible in principle until some physics (or rather biology) discovery explains how it would be possible. Until then, anything we engineer is an approximation as best.