an201.com AI Reflection

A newsletter on the philosophy, ethics & law of AI

The Mirror We
Didn’t Ask For

The Mirror We
Didn’t Ask For

Every technology changes what we do. This one is changing what we think we are.

This page is not about AI news, model releases, or productivity tips. It is about something quieter and more unsettling: the way artificial intelligence is forcing us to re-examine ideas we thought were settled — intelligence, expertise, evidence, even our own psychology. Here are three of the seeds this newsletter will grow from.

On intelligence

We never actually defined it

For a century, the best definition we had of human intelligence was embarrassingly circular: intelligence is whatever an intelligence test measures. We got away with it because nothing else could take the test.

Now something can. Machines built on pattern recognition — and on reasoning, which may be nothing more than pattern recognition chained together — perform so much like human minds that the old evasion no longer works. We are forced to ask what human intelligence actually is.

If recognizing emotion doesn’t require lived experience, what exactly did we think emotional intelligence was?

Even emotional intelligence, which felt safely human, has become uncertain: an AI can identify emotions with striking accuracy without ever having felt one.

On evidence

What counts as proof is changing

For decades, the gold standard of science was the randomized controlled trial: test a treatment on a population, average the results, declare an effect. Medicine, psychology, public policy — all built on it.

Personalized medicine is doing something fundamentally different. Instead of asking “does this work on average,” it asks how biology works in this one person, and targets the therapy accordingly. Randomized trials aren’t obsolete — but the very meaning of “an effective treatment” has shifted underneath us, from a statement about populations to a statement about individuals.

On making

The team is no longer the unit of creation

It used to be self-evident that building an application or a service required a team. That assumption is dissolving: AI can now do most of the work, with real limits remaining around cybersecurity and the judgment of an experienced engineer who understands what can go wrong.

But even those limits may prove temporary. When one person with a machine can do what once took a department, our ideas about work, expertise, and professional identity have to move too.

Why this exists

What this newsletter will do

These are only seeds. The questions they open — philosophical, ethical, legal — are the real subject: who we are, what our intelligence is and will become, how we define our head, our body, and everything around us.

I’ll explore them here, in plain language anybody can understand, because these questions no longer belong to specialists. They belong to all of us.

Read the next reflection

One essay at a time, when there is something worth saying. No noise.