Evergreen AI principles

Produced by DALL-E

”The stars move still, time runs, the clock will strike.” Marlowe, Dr. Faustus

Theory of general negativity

The AI hype cycle chunters along. Sure, “chunters” is a word. Search it up, as my kids say.

In the trough of disillusionment, general negativity is to be expected. I blame Einstein. I’m more pragmatic. The genie is out of the bottle and it’s not going away. So let’s get on with adapting to it.


I’m really enjoying Ethan Mollick’s book: Co-Intelligence: Living and Working with AI. He’s an associate professor at the Wharton School of the University of Pennsylvania, where he studies and teaches innovation and entrepreneurship. He also runs the excellent One Useful Thing newsletter. It’s on Substack, which I’m trying not to hold against him.

I recommend both. I’m not financially incentivised to.

Co-Intelligence by Ethan Mollick

One part of the book contains some evergreen principles for working with AI, even as the technology changes at breakneck speed. They’re useful and thought-provoking.

Principle 1: Always invite AI to the table

If you’re trying to get something done, try using AI. It’s the only way to truly grasp what’s possible. Sure, AI has weaknesses. What doesn’t? You may find that your weaknesses are mitigated by some of AI’s strengths. You may not. Just try it.

While doing this for myself, I’ve been repeatedly reminded of the concept of the connectome. A connectome is a comprehensive map of the neural connections in the brain. Due to the way LLMs use tokens rather than words, they often make connections between unexpected concepts. I think that’s a real strength, even if it can get surreal.

Principle 2: Be the human in the loop

Working with AI isn’t some sort of Faustian pact. You are responsible for the “craft” of working with these co-intelligences. Handing over agency to them completely, without critical thinking and review, is a recipe for all the negative outcomes you read about online.

Principle 3: Treat AI like a person (but tell it what kind of person it is)

Working with AI is easiest if you think of it as an alien person rather than a human-built machine. You have to guide and explain, set parameters, and have a two-way interaction.

Principle 4: Assume this is the worst AI you will ever use

Back to the hype cycle. Of course, there are flaws, ethical considerations, errors and so on. But this is a fast-moving field. The technology isn’t standing still. It’s a mistake to think that it’ll have no impact on your work or the trajectory of your organisation eventually, and maybe sooner than you think.

So, if you’re interested in the topic, for whatever reason, Ethan’s book is a tenner well spent.

Evergreen AI principles
Older post

Illuminating the dark forest

Could indieweb practices bring some illumination from the cozy web to the corporate dark forest?

Newer post

Getting started with automation

Want to reduce your workload and stress by automating some tasks? Find out how to get started.

Evergreen AI principles