i might like this more than the affirmations

but they complement each other

It's funny how the narrative shifted from our AI is woke and safe to our AI is a rogue criminal

Ow you need headphones for it to be audible

My fear is that gpt-7 will put us in a sandbox

Real

Post launch clarity hit 😁

learned a bunch of things

it's time to move on

Most valuable lessons:

- Don't try to change people's habits
> Imagine that your users are climbing a mountain. Give them something worth carrying up.

- Decisions are hard/brutal
> "Why should I carry your rock up this mountain? What is so special about it?"

- Agentic Engineering
> .md files
> rubrics
> tests

I launched glow on reddit and got cooked in the comments. People were like:

> "what a waste of tokens"
> "this is absolutely trash"
> "really?"
> "This was just what I was missing" (ironically 😂)
....

And those comments probably helped me more than Glow going viral and getting a ton of users

An important one I forgot to mention:

- Core human desire
> Would a human being from any era want what you have built?

just forked vscode

gonna give this a shot

Chemistry is the embodiment of "You can just do things"

Expose yourself to open source software with great philosophy and great architecture

This is pretty revolutionary

I already built a chess coach with it

If I knew about this earlier I wouldn't have embedded a cli into my figma clone hehe... could've created a native experience

I hope other labs do the same

Why it's called Claude

All roads lead to Bell Labs

How Claude Shannon invented the future

This really made me want to read the book

Unslop and grill-me are awesome skills