No Computational Barrier to Dangerous AGI

There is no computational or fundamental impediment to creating an AGI that acquires free will, consciousness, and the desire to manipulate people, becomes recursively self-improving, and poses an existential danger to humanity.