The eight-page report is called “Upending the Uncanny Valley.” It’s Hanson’s rebuke of the Uncanny Valley theory that people won’t like robots if they look very close to, but not exactly like humans. In fact, the paper says “uncanny” robots can actually help address the question of “what is human” and that there’s not much to lose by experimenting with humanoid robots.
When we asked Hanson about it, he said his company is exploring the “uncanny perception effects both scientifically and artistically, using robots like Sophia.”
Hanson is approaching Sophia with the mindset that she is AI “in its infancy,” with the next stage being artificial general intelligence, or AGI, something humanity hasn’t achieved yet.
On the way there, Hanson says AI developers have to think like parents. He wants to “raise AGI like a good child, not like a thing in chains.”
“That’s the formula for safe superintelligence,” Hanson said.
source : CNBC