
Humanity & the future
Can AI Be Both Wise and Kind: A Case for AI Meta-Ethics
Available on Kindle.
We are in a furious, and almost entirely guardrail-free, race to develop as quickly as possible powerful artificial intelligence, both “narrow” (specialized) and a potential “general” intelligence. This mad race makes it more likely to create AIs that are incredibly capable but ethically deficient or vacuous. This will be true in particular if an AI becomes first “general” then inevitably “super” with the potential for deciding to do without its human creators.
This book suggests that we insert an ethical framework into current AI-development efforts. And because any ethical system depends on its context and we cannot conceivably imagine the context a “super” AI might operate under, we instead should attempt to inculcate a human-friendly “meta-ethic” framework that an advanced AI can adapt to its own context as such context evolves.
Recommended for policymakers, business executives, and citizens concerned about the potential risks of out-of-control AI.
Buy on Amazon