I am, of course, referring to the latest wave of “AI” products like ChatGPT, Bard, and others. In the last 18 months, the world has been taken by surprise by how much progress has been made on the AI front, and computer software can now cover pretty well for flesh-and-bone software.
When people ask me if we have achieved artificial intelligence, I always answer, “I don’t know, and I don’t think it matters.” First, you should know I cannot know if your intelligence is like mine. I am certainly not smarter or better than you, but from my vantage point, I can only assess mine and, at best, how I react to yours. Maybe you make me feel cultivated, dumb or smart, but these are all reactions my brain created based on signals from you, not definitive proofs of your intelligence. Empathy and Occam’s razor make me conclude that you are very certainly as intelligent or more than I am, but these are all hypotheses that I have no real way of verifying. What if I showed similar empathy toward machines that provide me with signals that are aligned with yours?
On the other end, what I think is important and truly impactful as a human who is not necessarily creating these machines is what the AI can do, not how it does it. Who cares if it is intelligent or not if it can do the same or better than other “intelligent” beings?
And there, the last few years have taught us one thing: the machines are able to do pretty much everything we do, and better in many cases. It started with games like Chess and Go, but it’s clear that it’s also true for “creative” work or even treating us. By extrapolating, I’d say that there is nothing that a computer won’t be able to do better than us: calculating, playing, driving, researching, creating, deciding… etc.