As business —and the public at large— struggle to understand artificial intelligence and what it might become, I am free from any such anxiety. Arther C. Clark explained it back in 1962:

“Any sufficiently advanced technology is indistinguishable from magic.”

For now, you might say. Like all magic through the ages, we’ll figure out how the trick was done. And there are computer scientists today who write the AI code and understand how it works.

But if AI keeps “improving” at an exponential rate perhaps full understanding will remain just out of our grasp. And, yes, somebody wrote the code for today’s AI and understands how the trick is done. But who will know when AI is writing its own code?

For my part, I don’t really care how the trick is done (until it’s a bad trick, of course). For now I think about it the same way I think about “the hard problem” of consciousness. Nobody really understands what it is and how it comes to be. If some future AI achieves consciousness, and can explain it, I hope it doesn’t.