Andy Halliday

And it extended work into more hours of the day, often without the employees being asked to do that.

They just ended up doing that.

So on their own initiative, workers did more because AI made doing more possible and

made it more accessible.

And in many cases, it was intrinsically rewarding.

They had a positive experience of it, even though the productivity from the company's perspective must be improving dramatically, not just because the AI is doing a good bit of the work, but because the people are engaged and motivated by working with AI.

You're on mute, Beth.

Right.

This reminds me of something that Carl talked about when the first agentic browsers came out and you could launch multiple tabs.

So you have various tasks on multiple projects spreading out now across a browser window, an agentic browser like Comet or Atlas or now the Cloud Code extension into Chrome.

Like you can do the same kind of thing.

So you can...

Spin up a little agent, have it go off and do something.

And of course, the agents are smart to not just continue autonomously to some unobserved conclusion, but they're going to want interaction with the human.

So now you have this emerging possibility that you have, like on my desktop, there have to be presently, there's probably 80 tabs open.

Well, what if 20 of those are different agents that are waiting for my response?

And now I'm trying to multitask across multiple projects with multiple agents spun out and going off in different directions.

That's overwhelming.

And there's value to maintaining flow and focus in one endeavor rather

You know, rather than having to split your attention and attempt what is not a native capability of human cognition, and that is multitasking.