My interview with Dr. Walid Saba.

And a short summary of it.

‍

It seems that for once, Europe is leading in technology discussions as they are currently working on an AI policy. However, a group of 1000+ mainly US individuals signed a paper last week asking to halt further development beyond ChatGPT-4. Just before the release of this letter, I had an interview with Dr. Walid Saba, a 30-year veteran in AI and language research, about ChatGPT and its claims to lead to AI general intelligence (AGI) and generative capabilities. You can see a summary of this 50-minute interview or watch it in full length here.

Regarding this discussion, first of all, setting boundaries for the use of any technology is a good thing. Does it have to be very complicated? No. The harm of technology is nothing new, and people should be able to agree on common boundaries that are legally enforceable. Banning the entire development, however, is not something I agree with, and here are my reasons.

Only countries like China, Russia, Iran, North Korea, and - maybe not surprisingly - Italy have banned ChatGPT. It is often these countries that do not want their people to use any information technology so they can manipulate public opinions and secure their illegitimate governments.

The paper was signed by individuals such as Elon Musk, who is very obviously annoyed that his $100M investment in Open AI was cashed in by others while he has not received anything. I do not trust Elon to be quite frank, and I am quite sure he is working on an alternative. So to sign this letter is helping his own behind-the-scenes activities. Elon Musk is purely interested in it from a commercial standpoint, and I do not buy his other reasons. Someone who reinstated a criminal president on Twitter should not be listened to any longer.

As Walid Saba pointed out, there is no threat at all by ChatGPT to take over the world in an evil way. Therefore, to halt all development for six months, as suggested, seems quite funny, especially since the letter was signed by individuals who theoretically understand AI quite well.

‍

picture source: https://www.nytimes.com/2022/04/15/magazine/ai-language.html