I didn't comment on your Feynman link as I have watched all that is available on him on youtube many times as I am a big Feynman fan.

I am glad you enjoyed our exchange.

Did you read this entry on my link:
In 2018, Google Duplex voice AI called a hairdresser and successfully made an appointment in front of the audience. The hairdresser did not recognize she was speaking to an AI. Considered to be a groundbreaking achievement in AI voice technology, Google Duplex is also far from passing the Turing test.

Duplex is a deep learning system representing the ‘Second Wave of AI’ – trained with hundreds of hours at performing very narrow tasks. Real-time learning, deep understanding, reasoning requires true cognitive abilities that none of the Second Wave AI programs have. As soon as the human would lead the conversation in a different direction, Google Duplex would fail.

I think it's a direct criticism of the vid you posted, is it not?

Your quote from Hassabis is partly why the turing test has not been passed yet by any AI system.
Surely you would withdraw your 'passed it with flying colours' claim!

I fully admit that I don't know the details on the workings of the most up-to-date AI systems involving artificial neural nets but you would need to source me some very convincing scientific evidence that current experts in the field don't know how current artificial neural nets produce the results they do based on your typing of:

Even now we don't understand what artificial neural networks are doing when they give their responses, it's too complex. — punos

I think it's a direct criticism of the vid you posted, is it not? — universeness

Yes, you are right that AI has not reached that level yet, and i probably spoke too soon about the "flying colors", but my point is more in the direction that if AI can already fool some of the people some of the times, then it probably only means it won't be too long before it can fool most of the people most of the time. It's like video games and graphics when in the early days the quality and resolution was ultra low, while today we have near realistic renderings. Elon Musk has made this point. Some of those graphics fool some people into thinking they are real images, and it will only get better. Eventually no one would be able to tell the difference, consider AI deep fakes.

I'm subscribed to this channel which helps keep me up to date on AI graphics developments. It's called

Two Minute Papers:
https://www.youtube.com/c/K%C3%A1rolyZsolnai

I fully admit that I don't know the details on the workings of the most up-to-date AI systems involving artificial neural nets but you would need to source me some very convincing scientific evidence that current experts in the field don't know how current artificial neural nets produce the results they do based on your typing of: — universeness

This article i think can help clarify what i mean:
https://towardsdatascience.com/why-we-will-never-open-deep-learnings-black-box-4c27cd335118

I appreciate the links. I will have a look!
Again, thanks for the very interesting exchange :smile: I am surprised that your line of thought did not attract more contributors from the TPF population.
Maybe I speak too soon.

You contributed :smile: ,we can speak any time i'm on, it was a pleasure. :up: