Yann Lecun once liked one of my comments on Facebook. I didn't know who he was at the time and can't remember what I said but the name looked interesting. But Ben Goertzel, the one who created Sophia and is into AGI, is a facebook friend of mine. Not that I have ever met him in person, or even spoken to him and he didn't even respond to my PM asking how he found me when he sent me a friend request. He seems to send friend requests to any published researcher in AI and so conversations sometimes occur.
People are attributing far too much intelligence to Sophia. But on the other hand I will still seriously listen to anyone who actually creates AI over someone like Nick Bostrom, Ray Kurzweil or Kevin Warwick. (I once brought down a shit storm on myself for slating Ray Kurzweil on facebook, there are a hell of a lot of AI groupies out there). Thing is, you can think about AI as much as you want. but you can only make progress by actually trying to create AI, seeing what works and what doesn't. This is because it is the only way to find out what the real challenges are. Creating a physical body that can act like a human, even if it is totally scripted and has very limited intelligence is still a worthwhile endeavour.
Most AI is trickery which looks more impressive than it is. That's not to say that we don't learn anything from doing it, but the challenge with AI is making it scale up.
People are attributing far too much intelligence to Sophia. But on the other hand I will still seriously listen to anyone who actually creates AI over someone like Nick Bostrom, Ray Kurzweil or Kevin Warwick. (I once brought down a shit storm on myself for slating Ray Kurzweil on facebook, there are a hell of a lot of AI groupies out there). Thing is, you can think about AI as much as you want. but you can only make progress by actually trying to create AI, seeing what works and what doesn't. This is because it is the only way to find out what the real challenges are. Creating a physical body that can act like a human, even if it is totally scripted and has very limited intelligence is still a worthwhile endeavour.
Most AI is trickery which looks more impressive than it is. That's not to say that we don't learn anything from doing it, but the challenge with AI is making it scale up.
