AI and cognitive decline

Last week the Journal published a story about research showing that cabbies and ambulance drivers had low incidence of Alzheimer's Disease, presumably because they are using their brains dynamically much of the time to problem-solve on the fly and figure out how to get from place to place quickly under changing conditions. This basically accords with my thesis about the deracinative effects of using Google Maps all the time, how one's ability to remember directions and think spatially is gradually eroded by the disappearance of a need to do so.

So what the heck do we think AI is going to do? There is broad speculation about how AI will help us solve ever more complicated problems and think more dynamically, blah blah blah. I for one ain't buying it. I think it's just like muscles, you lose them or lose them.

I do basic math all day even when I don't have to. The Schwab trading interface has a handy little calculator to figure out how many shares of, say, Gamestop one can buy for $10,000. I don't use it. I look at the price and then estimate the number of shares and then refine my guessing as I go till I'm close enough. The point is not to have the math done, but to be able to do the math.

I could use an AI to write blog posts. But what would be the point of that? I'm not trying to produce text but to do the thinking that allows me to write something interesting. To keep thinking and think better. I'm sure there are ways that I could partner with an AI fruitfully, eventually I may go there. I just listened to Ethan Mollick's Cointelligence in the car after seeing it pop up on a bunch of best of the year lists. Good book, smart guy. He has learned to use ChatGPT in interesting ways and he challenges his students to do the same, to the point of mandating that they use it. I'm just not feeling it. I feel a deep Andy Rooney thing coming on.

Maybe in 2025. Talk to you then.