Cal Newport

All we had was coding agents.

That's the only thing that we worked on that whole year.

It was supposed to, I mean, I have the receipts early 2025, all of these executives saying your work as a knowledge worker, not as a computer programmer, but just as a knowledge worker is going to be largely done with agents.

You're going to have agents are going to be a major part of your workforce and

in just a normal office setting.

And none of that happened because it turns out just asking an LLM, give me a plan for doing X, doesn't often actually produce a workable plan.

And to what end?

That's the real question, right?

I mean, I've seen these proposals.

I mean, basically where a lot of these proposals go, I mean, it's the agents were supposed to, we thought that we could just make AI do anything.

So we'll just, we'll have it use the mouse and just use our computers for us.

Oh, that's hard.

We don't know how to do that.

All right.

So what we'll do is we'll rewire all applications that anyone uses in the internet so that we don't actually have to use the mouse.

It can have a text interface.

So that an LLM, like they do, the coding agents do, can give a description of how to do something in Excel in text without having to actually move a mouse or click things around.

And then these evolve to say, okay, well, what's the one type of instruction that we're good at producing?

Because they get...

When LLMs produce plans, they're directionally correct plans.