• Log In
  • Sign Up
    • Not to take anything away from your awesome chatbot experiment, but I feel a little like Crocodile Dundee when he said, That’s not a knife. This is a knife.”

      Here is IBM’s Project Debater taking on one of the greatest debaters on earth:

      They battled to a near draw on a tough subject. My take is she vastly outdid the human in knowledge, but he said he was able to pull out the win by using personal experience, which she didn’t have, and appealing to human emotion.

      I have the deepest respect for both Ben & Richard and their logic, but having been involved in AI for so many years, I no longer bet against anything. We thought for years that it could never do pattern recognition like humans and we used as an example a line drawing of Lincoln’s profile, which every American could identify instantly and no computer had a chance at. Now, AI crushes humans at some facial recognition tasks along with recognizing tumors on MRIs when competing against trained radiologists.

      So far human + AI > AI or human alone in most cases, and it isn’t yet very good with humor or emotion, but I’m not betting against it.

    • I just finished watching the whole debate and that was fascinating! The AI did an amazing job at gathering information and then synthesizing it in a coherent and argumentative way such that it could hold its own against a world class debater. I think the key that they said, and you said is that AI+humans has the potential to do a lot of good and accomplish a lot.

      When I was learning about AI in my philosophy classes, both at Foothill College and UC Berkeley, we often viewed AI through the prism of whether or not it could have consciousness in the same way that we do. That our brains are computers and thus if we create a sophisticated enough AI brain, that too could have consciousness in the same way that we do.

      Upon watching the debate, my opinion about whether or not AI could develop consciousness like humans didn't change. I don't think more algorithms or better computer science will be able to create AI that is able to be self-aware of its own existence like we are. In other words, I think the age old concern of AI replacing us, taking over the world, and becoming out of our own control is just science fiction. I don't foresee that ever happening.

      That said, I think what did change for me upon watching that debate is a better understanding of the purpose of AI and how it can help enrich our lives. Rather than being afraid that AI will overpower us and become this unstoppable force that becomes more intelligent and self-aware than us, we need to embrace AI and see how far we can take it. The further we can take AI, the more we can push it to be better, the more our lives can be enriched. It's not AI vs. human that the future is heading towards, but rather AI+ human. That should excite all of us.

    • Man you keep me busy engaging with fascinating content! LOL. I'll give this a listen on my drive home. I am a big fan of Intelligence Squared and missed this on my feed so thank you for this.

    • I dunno, Ben. As I get blown away by each advance of AI and the unexpected ways it thinks, I inch closer to where Elon Musk is. He thinks AI is more dangerous than nuclear weapons.

      I think presidents Putin and Xi really get this and are focused on it, while Britain dithers over Brexit and us over a border wall.

    • I guess I didn’t really think of that angle, but yes, you are absolutely correct that AI could be used for very devious and cunning purposes. No question about it. But, what would be behind such purposes are evil humans. More broadly speaking the same can be said for technology, which AI is a branch of. Technology can be used as a force for good, but in the hands of the wrong people can become a weapon for evil. We certainly live in both exciting and scary times in that vein.

    • The debate with the AI bot was very interesting. AI was much more informative than the human debater but definitely fell a little short on critical thinking. One of the developers who was interviewed said it the best when he mentioned the tech was more for assisting human awareness than engaging another in a debate. Just the assist in human reasoning alone could be a significant development. With the proper human controls of course. When AI takes over the controls is when we are screwed. Very impressive though.!!!