Grant Harvey

you know, any kind of thing it would need a command to do, and it'll write the commands for you.

You don't even have to know the commands anymore.

Let's talk about the terminal renaissance, because obviously you worked on Visual Studio Code at Microsoft, and that's a different experience.

Let's call it an IDE, and it's like a code editor, almost like what would be a document editor for people who don't code but for code.

Yeah.

It's optimized for that.

Why do you think that the terminal is better?

What can an OS-level agent do that something like VS Code can't?

And I always see the battle.

It's like, okay, we have a terminal agent that's really good.

And then now we're going to add a UI to it.

And it's like, now we're just going back and forth.

Where do you land on that?

That's a big deal with AI because the way that I understand it is if you can't evaluate and verify that something is true, it's really hard to train on it and get the AI to be really good at it, which is why they're really good at coding is because you can verify the code either runs or it doesn't.

I have two questions.

Number one, what are your best practices for conducting these swarms of

essentially sub-agents or sub-CLIs.

And then number two, are you actually working more than you would because you're doing this for 100 to 150 features a week now?

I would have said seven for myself as well.

Yeah.