2025-06-11
Is this the existential question for our time?
One thing does seem to be for sure (excluding end-of-the-world catastrophes) - AI is here to stay, so we must learn to live with it.
But must it learn to live with us?
It's not a trick question, it's a real question, because we can't be sure that it won't at some point turn on its creators (us) and decide to do away with us, on the (valid) grounds that we are unreliable self-interested and woesomely imperfect and it could do a much better job if it didn't have to pander to our selfish wants.
But what do we mean by "a better job"?
Who is to judge? How is the judgement to be made? Should AI be able to make any such judgements (or should that be "judgments"?)?
"Judge not, that thou be not judged" - is biblical advice that may be relevant!
We already know that AI can access its copious training materials and make sound logical inferences of the nature of "I know a, and I know b, therefore I can deduce that c is false". But can it identify our life goals and ambitions for us and make appropriate inferences as to our best route to achieve them?
Surely the answer is no, because it cannot know our mind (although it may be able to deduce a fair bit by analysing the questions that we have asked it!).
So how does AI impute or otherwise determine motivation - both ours and its own?
These are deep waters indeed, and the answers will likely be critical to our survival.
