On Friday, June 13, 2025 at 9:44:41 PM UTC-5 tbp1...@gmail.com wrote: I have read Turing's paper and I think it's something of a masterpiece. He finds a way to avoid figuring out what "thinking" is and still make progress. We have now had 75 more years of experience than Turing had in 1950, and it's no wonder if there were things he hadn't figured out. And we still don't have a good handle on what "thinking" is. His original proposed game is more sophisticated than what is usually meant nowadays when people talk about "the Turing test", and we have have had more chances to see where these tests have turned out not to be perfect. And after all, his ideas seem to have been evolving in the several years after the paper was published. Were he still alive we can be sure they would have evolved further.
There's no point is arguing whether Turing got everything right. His ideas and presentation opened up people's minds in the area of thinking machines, and even now he doesn't seem to have been far from the mark. Thanks, Thomas. Your views are my own. Let's recall how stunning the recent progress in AI is. Ten years ago, nobody, not Turing, not Geoffrey Hinton <https://en.wikipedia.org/wiki/Geoffrey_Hinton>, not Demis Hassabis <https://en.wikipedia.org/wiki/Demis_Hassabis>, not anyone else, could have predicted the effectiveness of LLM's using attention-based architectures. This progress has changed all of our *questions*. Alan Turing surely would have something profound to say if he were alive today. Specifically, in 2016 (almost ten years ago!!) AlphaGo defeated Leo Sedol <https://en.wikipedia.org/wiki/AlphaGo_versus_Lee_Sedol>, who was one of the world's top Go players at the time. I clearly recall thinking this was the most shocking scientific result in my entire life. It still is. But earlier results had already showed that neural nets have incredible power. In 2015 (only *one* year earlier!!!) Deepmind had mastered Atari video games. This two-minute papers video <https://www.youtube.com/watch?v=dJ4rWhpAGFI> gives the details. I remember seeing a video something like this one <https://www.youtube.com/watch?v=PH1Ruc3TPqE>. It's incredible. But I though Go was still out of reach. How wrong I was. And then came AlphaZero <https://en.wikipedia.org/wiki/AlphaZero>, AlphaFold <https://en.wikipedia.org/wiki/AlphaFold>, and AlphaEvolve <https://en.wikipedia.org/wiki/AlphaEvolve>. The last two projects solve problems that are vastly more complex than Go. *Summary* It's easy to forget just how shocking the last 10 year's progress has been. Scientific life before 2015 is fundamentally different from today's. Edward -- You received this message because you are subscribed to the Google Groups "leo-editor" group. To unsubscribe from this group and stop receiving emails from it, send an email to leo-editor+unsubscr...@googlegroups.com. To view this discussion visit https://groups.google.com/d/msgid/leo-editor/4bf961ac-debc-4167-bb61-f2c16f63b4e4n%40googlegroups.com.
