Nathaniel Whittemore

which is all knowledge workers.

This inflection point, which really happened around the end of last year and the beginning of this year, basically between November 25 and January 26, shift the advantage gap into overdrive.

The people using agents are seeing compounding value while the people using regular chat continue to see linear gains only.

Now, when it comes to the business model side, it is absolutely the case that the people who are using agents are spending far more money than those who are using regular chat.

The difference between seat-based pricing and usage-based pricing is the difference between the $3 billion run rate that Anthropic had last year and the $47 billion run rate that they're currently on.

If you've been listening to this show at all over the last few weeks, the number one most dominant and most important theme has been the shift from the token subsidy era to the token scarcity era, where the business models are all shifting to sell people the tokens that they're actually consuming with lots and lots of consequent changes.

What I think though is a mistake in just assuming that this is a business model question and an IPO question is to think the thing that primarily these companies care about is the revenue scoreboard.

That obviously matters, but the reason that I think you're going to see a major change in the interfaces and user experiences that OpenAI and Anthropic put in front of their customers is a recognition of the fact primarily that the people using agents are getting more value and a desire to use interfaces and user experiences to bring more of that to everyone else.

If you are watching closely, there is even a gap between the power users and the power power users, reflecting just how quickly user experience patterns are evolving.

OpenClock creator and now OpenAI employee Peter Steinberger wrote, Here's your monthly reminder that you shouldn't be prompting coding agents anymore.

You should be designing loops that prompt your agents.

On CNBC recently, Claude code creator Boris Cherny told the host that about six months ago, he shifted from writing code by hand to prompting Claude to write all of his code, but it turns out that even that isn't the limit of how things are changing.

Here's a snippet of a conversation with Boris from just last week.

Now this idea of loops isn't totally new.

Last year and into the beginning of this year, we had a lot of discussion around the Ralph Wiggum loop, which is basically a way to set up your coding agents so they would continuously try and try again without you having to sit there and prompt and interact every time they came up against the challenge.

Then in March we had Andrej Karpathy talking about auto-research, which was a specific type of loop designed to improve an actual AI model.

And now we've got the slash-goal primitive, which is a way that both OpenAI and Anthropic have embedded the idea of loops into the core experience of their major coding tools, Cloud Code and Codex.

The goal in each of these cases is to require less human intervention and get the AI to run for longer and longer, being able to fix its own mistakes, improve its own results, and ultimately accomplish much more comprehensive and complex tasks.

And so the point is, if you've already got people who are using agents opening up a significant advantage gap compared to people who are just using regular chat, and then you've got the vanguard of people at the labs who are even going a step farther in terms of how they're getting the most value, it strikes me as obvious that one of the places that the labs can try to democratize those experiences to more users is via the user experiences of their core app interfaces.

And that, I think, is the key point of the ChatGPT overhaul.