Dan Shipper

Yeah, I think that that's a reasonable frame.

And I wish that they would say that more because it gives...

It's just such a bad look for all of technology for CEOs to be doing that.

Did you see the Standard Chartered CEO got in hot water?

Yeah, and I think this all changes, and the lines between those terms are all going to change because they're all sort of relative to the capabilities of technology, and that's all changing too.

But I can tell you that what we find internally at Every, and again, we're one of those places, if you add someone in Slack, it's like toss up whether or not it's a bot.

And everyone's using Codex and Cloud Code to do all their work every day in the engineering organ, like for writing, for editing, for design, all that kind of stuff.

And I can tell you that every agent needs a human.

The further away an agent is from a human who's managing it, the worse it does.

And you see this, too, even in the scaled AI companies like inside of OpenAI or inside of Anthropic.

they do have company-wide bots that you can add, but they're run by teams of people.

And I think that that's actually a really interesting and pretty stable phenomena based on how these agents work.

Obviously, they do more and more complex work, but what we see is that

Someone on our team, Kieran Klassen, calls it the human sandwich.

AI collapses tasks that used to take hours into a few minutes, but the human is still kind of like the sandwich on either end or the bookend on either end who's framing the task or evaluating it when it's done.

And what's really interesting is even though it can do

even though AI can do expert human work, my experience internally, and I think you find this across the AI industry, is it actually, even though AI can do expert human work, it actually increases the demand for human experts.

Because what happens is,

You can get expert human work out of an AI, you can get pretty good writing or pretty good images or pretty good code, but it's all based on yesterday's competence.

It's all based on what is in the training data from yesterday.