Timeline

I have never had patience for philosophy.

It always felt like stacking words on top of words until the chain of meaning was deep enough to forgot it doesn't connect to anything. Great fun for people who like arguing about definitions, but otherwise not interesting.

Here's a shocker. I am an engineer. A software engineer. Not a phenomenal software engineer, but I’ve worked at Amazon, Uber, was CTO of a startup, and am currently at OpenAI.

In 3 years, LLMs has grown from an annoying autocomplete to Astra and Opus 5.5, are the first models that are better than me at coding—better than me at the thing that I am best at.

If the next generation of models are not better than everyone at everything, then it will be the version after that. It seems that humans will be left to develop the only thing AI won’t be able to… our selves.

All we have left is "How can I be a good person?"

All we have left, is philosophy.

Fortunately, philosophy has only ever been bad at answering questions. As a means for asking them, it is unmatched.

And so we begin...