How can we increase the likelihood that AGI will enhance, rather than inhibit, social coherence by 2035?
That’s the question which the new Real-Time Delphi survey by the Millennium Project seeks to answer.
I drew attention to that survey in the latest issue of the London Futurists newsletter. (See item #7 in that newsletter.)
Prof Bruce Lloyd, a long-time supporter of both the Millennium Project and London Futurists, replied to me with a question:
Rather than just asking humans to take part in this survey, shouldn’t the world’s best AI systems also be engaged in the process?
To demonstrate what a set of AIs might offer in response to the initial question, Bruce sent me a set of 13 slides, produced by Gemini’s Notebook after reviewing the output of 10 different AI models that each have a claim, in their own way, to be considered among the world’s “most advanced” AI systems.
I’ve embedded that PDF below. (Click on it, then use the arrow buttons to view the different slides it contains.)
What do you think about it?
Whether or not you agree with every claim on each of these slides, I think you’ll find it a stimulating exercise to consider the different ideas offered.
And once you’ve done that, you’ll be better placed to head toward the Millennium Project survey (there’s a link to it in the London Futurists newsletter, as mentioned above) and offer your own human opinions!