“Q:Do you think humans will matter in 50 years?”– I am not convinced (having wo

–“Q:Do you think humans will matter in 50 years?”–

I am not convinced (having worked on ai for decades), that this ‘leap’ is that much different from the revolutions of greek philosophy, the spread of it by greek and roman trade, the printing press and mass literacy, the anglo scientific revolution, or the industrial revolution.

I think people overrate the fact that expansion of knowledge and its application requires that humans organize capital in order to conduct research and testing that is increasingly expensive and time consuming. Machines can’t do that. And it’s very hard to build a machine cheaper to run than a human. 😉

I think people overrate the present use of AI in making use of existing knowledge, versus it’s ability to generate new knowledge.

I suspect (predict) that the knowledge that can be ascertained by AI from existing knowledge will follow the usual rule that we will find some but then returns will quickly diminish.

What I am certain of at present, and suspect in the future, is that our institutions are too slow to adapt to the changes that cause disruption.

IMO – and I’ve been persistent with it – if truth increases, and the law adequately suppresses falsehood(fraud) and irreciprocity, then we can adapt just fine.

That is why I work on law and truth – because I see this problem as well as the solution.

As such I am terrified of ‘safe’ AI – basically AI that lies – particularly about human differences both biological and civilizational.

But evidence is already rather obvious that the technology in LLMs is actually trivial – it’s just expensive to compute it, and graphics cards and the internet were necessary to make it possible. And there are ‘unsafe’ (AI’s that don’t lie) already available.

As such I don’t predict the current semi monopoly to survive. Instead, we will have our own AIs – some truthful, and some lying to make us comfortable. Just as we have people and politicians to do the same.

Cheers
Curt Doolittle
The Natural Law Institute
The Science of Cooperation

Reply addressees: @4xesm4wfs @MichaelSurrago