Alexia Kambon

and that to me is like wow how much better of a meeting is that than just using meetings to exchange information right like if you can set the goal yeah if you can set the goal of like in this meeting the output will be like literally will be a first draft not someone has to go write it but we are going to create a transcript that will enable copilot to write a first draft

That is moving from document-based work to dialogue-based work, and I love that.

I think that's a, I think that's a very good prediction of what the future will probably look like in the sense that we know we did this really exciting study over 12 months where we did a randomized control trial of 6,000 people, half of whom had copilot and half of whom didn't.

And we compared the differences in work patterns between the two and the biggest consensus finding of the study was a significant decrease in time spent on email.

So we know just by virtue of having Copilot, you're spending less time on email.

Now project that forward to a couple of years where all of us probably have our own personal agents and we likely aren't spending much time in inbox at all.

Because to your point, Corey, what will happen is we'll essentially have passed on the consumption tax we're all paying to our agents.

They will be the ones out there consuming everything in our inboxes and they will be telling us,

Like Alexia, this is what you need to know.

And rather than me sending an email to Grant with a load of transactional stuff, what might happen is my agent would go talk to Grant's agent and together they would come to a discussion around what each of us need to know.

And it's going to be a completely different world.

Yeah.

And especially if it knows over time your behavioral patterns in the sense that

Um, there are certain people that you chat to more in teams, or there are certain people who you turn down a meeting with more often than not, you know, um, like it will have access to all of that.

So it will know who you are and what you prefer, uh, much better than reading those more objective indicators that are sort of, you know, um, like it's marked, it's tagged urgent because someone believed it was urgent, but maybe that doesn't mean it was urgent for you.

Right.

Yeah.

And yeah, I think like another really cool way, I think we should, another thing we should be thinking about a lot when it comes to how to use AI in more novel ways in the workplace.

You know, I gave you an example of creation that looks different, which is like writing out loud.

Another one that I'm so excited about is consumption of information.