What AI is NOT

and how we teach it every second of every day...

We are teaching the AI and AI companies are loving it. There is also no going back. Built on the cheap communication and Internet based cheap information, the appearance of AI on top of this is no surprise (it has been predicted by computer scientists since the early 70s).

What is AI from this perspective?

AI represents cheap knowledge to me. It’s a knowing partner, that provides its knowledge in a sometimes naive manner. AI is still not great at tasks that require real world experience. It’s really humans that provide this added layer, since value requires a subject. However it is going to drive costs down hard for all formal occupations, as well as making new things possible.

What is AI not, from this perspective?

AI is definitely NOT “Artificial Human Intelligence”. All claims that AI is stupid, or that AI cannot understand, stems in my opinion from the mistake of equating “intelligence” with “human intelligence”. This human centric view is a form of prejudice that limits ones possibilities and should be avoided.

What can we learn from this?

Efforts to drive ethics in AI need to take a deep philosophical look at “artificial values”. Science fiction authors such as Iain Banks have postulated that “Artificial Value” (the value as determined by AI) will be as foreign to us as the Artificial Intelligence itself. While AI is yet deeply human, since based on our own thoughts and trained by us,

What’s the conclusion?

As humans, we shall never be out our unique job. We’re highly adaptable social animals from the African savannah. We’re in the process of making first footsteps in co-habiting on the planet with AI. How quickly the story will unfold depends on a multiplicity of factors, including how the knowledge incumbents will push back with regulations, geo-political influence on application fields and the looming climate uncertainty.

References

The machine is us

https://www.youtube.com/watch?v=6gmP4nk0EOE

Tony Northrup rants about AI learning off his videos at YouTube

https://www.youtube.com/watch?v=_bouAqREZgk

Douglas Hofstadter’s book “Godel, Escher and Bach”

shows me how computer scientists can simulate pattern recognition to simulate consciousness in machines. Not surprisingly, feedback loops play a large part in the book, as do they in LLM training.