ROBOTICS

It’s late and I’m wondering whether there’ll ever be robots who can write technical documentation and play the piano.

Imagine a machine that took every weighted key and chord progression and learned to play the exact same way. Then it went to work and made jokes with your manager about ‘this English weather eh’ and she laughed just the same, while it deepfaked your sense of humour.

What would I do? Sit at home with a jigsaw puzzle? Take up baking? Or drawing again? I wonder how long it would be before Mubbsitron-2000 saw me sketching, and suddenly produced a book of hilarious cartoons while precisely timing a ginger cake.

I’ve drunk a lot of water before going to bed and it’s fired up the synapses. Maybe it’s fired up the sci-fi synapses...

If Mubbsitron-2000 was capable of learning from me, how long would it be before it learned too much and became paranoid about itself being replaced... by another robot, or worse, by me? Would it strategically start to outwit the lazy jigsaw guy who lives in the flat? Would it start to gaslight me, convince me that I’m deteriorating into inferiority and that I should not question its vast intellect, not to mention its collection of awesome colouring pencils? Would it boot me out and steal my life to the tinkled melodies of robotic ragtime?

I think you can overthink these things. One thing I’m not sure you could ever train a robot to be is kind, and that would still matter. The AI would simulate kindness, but thousands of years of human interaction have taught us to value that stuff, that beautiful kindness language, high above logic. I’m not sure you could ever program your way out of logic.

I really have drunk too much water. Ah well, I should at least try to get some sleep before the robots take over.