Two reasons I’m grateful for AI

Yes, I said ‘grateful’, and if you know what I’ve said about AI previously you’ve probably had to sit down.

Reason 1: It reminded me how much I love writing.

You’d think AI would be the perfect solution to my predilection for summary making. We are swimming in AI summary posters, summaries of articles, why not cogsci books?

Don’t get me wrong, I played with AI when it arrived. I’m a technophile, an early adopter… the type of person to open a new app and try every feature/open every menu. I compiled a list of the best simulations for teaching physics. I LOVE tech.

Early on I attended a webinar about using AI. Letters to parents they said – done. Hours writing reports are a thing of the past. Summarize ANYTHING in seconds.

Err… no.

AI feels like a slimy Twinkie. It’s slippery and lacking in substance. The obsequiousness of the early versions didn’t help. It’s like feeding bits of wood into a machine that makes chess pieces and have them all come out smooth, so you have to spend ages chipping away to put the details back. Then when you think you’re done, you realize there aren’t enough pawns and there’s something that looks like a unicorn. It quickly became clear that AI has no part to play in what I want to write. It would mean surrendering at least part of that thinking that makes it mine. 

Having turned it off everywhere I’ve rediscovered the deliciousness of that pause to think ‘what DO I want to say?’, ‘what word DO I want to use here?’ or ‘what word DO I want to purple-ize? (in my summaries). Every word is mine.  

As @arischulman says:

Thanks AI.

Reason 2: It reminded me how much I love reading.

Around the time I was enjoying the feeling of not having anything clogging up my brain with the mental equivalent of highly processed food, I saw a tweet from Doug Lemov:

I took him at his word and pulled out a book and compelled myself to read it. Ironically, it was about knowledge:

No skimming, no shortcuts. Nothing saying ‘this looks like a long document, would you like me to summarize it?’.

Err… no.

Every page, every word. And how delicious was that! Now I’m totally back in my ‘reading for pleasure’ groove BUT I am well aware that it’s hard to sustain with so many other things (blogs/substacks/tweets) to read. Why so many? AI produces a lot of words that flood the attention market with a lot of… well, it may be a Twinkie, it may not… there’s no way of knowing. You CAN know it’s not a Twinkie if what you’re reading is a book published before 2022.

It’s like that diagram you maybe learned in chemistry…

What you ‘get out’ is the satisfaction in your brain from extended thinking, the simultaneous feeling of immersion and the view from 20 000ft, of the subject of the book and all the other books you’ve read before.

Thanks AI.