Off the Record

It was a one-day retreat, early morning to late afternoon, the kind of gathering built to make founders disconnect, unplug, and sit in a room with other people actually living through the thing.

I arrived fractured. My mind was on a product I was about to launch and on coordinating with my team, and I could feel the pull of it all morning even as I tried to stay present. The conversations were good and the people were generous, and still I caught myself checking the corner of my mind where the to-do list lived, and I am sure I was not the only one doing it.

By late afternoon the schedule had relaxed, the way it does at good retreats, into the kind of conversation that only happens when the recording stops. Phones were in pockets, nobody was livestreaming, and somebody had just poured a second round.

The conversation had been circling the same few topics all day: the future of work, whether SaaS as we knew it would still exist in three years, the vibe coding apps being spun up in a weekend and shipped before anyone could ask whether they should exist, and the quiet gaps in the enterprises people were selling into, where the budget was real but the strategy was not.

And then someone, I won’t say who, said what everyone in the room was already thinking.

“I don’t know if I can keep up anymore.”

Nobody rushed to fill the silence, which was the tell.

I have been in a lot of rooms this year: rooms full of executives announcing their AI transformations, rooms full of keynote speakers quoting frontier lab research, rooms full of people on stages saying the word “exponential” as if the speed itself were a badge. And then there are rooms like this one.

In the first kind of room, everyone is confident: decks are polished, the future is bright, adoption curves bend up and to the right. In the second kind of room, the confidence falls away. What replaces it is not despair but something quieter and harder to name. A shared recognition that the pace of this has become almost unliveable, and that most people are performing a version of themselves who can keep up, even when privately they suspect they cannot.

The gap between those two rooms is not a communications problem. It is the defining emotional condition of this moment.

The people building are running the fastest.

A founder shipping an AI-native product right now is not just racing competitors, they are racing the foundation models themselves. The ground shifts every few weeks, and a single model release from one of the frontier labs can make half a roadmap irrelevant overnight. There is a genuine question, one that does not get asked in the fundraising deck, about whether the thing they are building will still make sense by the time they finish building it.

At the retreat I listened to a founder describe the third pivot of his company this year. It was late enough in the day that he was no longer pitching anyone, just thinking out loud. He described it not in the triumphant language of iteration but in the exhausted language of someone who has rebuilt his life around a moving target three times and is beginning to suspect the target is not the only thing moving. The SaaS model he had grown up on was being questioned openly, the pricing assumptions were collapsing, and the customer expectations he was building toward were being shaped by a handful of consumer products that set a bar no B2B team could realistically meet.

He said, almost offhand, “I am not sure my company still has a reason to exist by the end of the year,” and then he laughed, because that is what people do at a retreat. He was not actually joking, and nobody in the circle pretended he was.

The people using these tools are running just as fast, in a different way.

I have spent a large portion of my career training end users on new technology, more than 17,000 of them on Copilot alone, and I know what it looks like when someone sits down in front of a tool they did not ask for, do not understand, and are being quietly measured against. What I see in their faces is more often a very specific kind of fear.

They are being told, in every news cycle and every town hall, that the technology is here and remarkable, that it will make them more productive, and that the ones who adopt fastest will win. What they are not being told honestly is that nobody fully knows what “adopt” means inside their particular workflow, that the quoted productivity gains are averages hiding enormous variance, and that the measurement systems evaluating their output are themselves in flux.

So they perform. They post about their AI-assisted weekend, quote prompt techniques on LinkedIn, sign up for courses, and watch the newsletters stack up unread because the anxiety of not reading them is lower than the anxiety of trying to catch up on all of them at once.

The private version of their experience is a low, persistent hum of falling behind.

The founder and the end user would probably not sit down to a meal and recognize themselves in each other. One is scaling a company while the other is being scaled by one; one is chasing product-market fit while the other is chasing the output metric of a manager who changed it last month. And yet they are both living inside the same internal system.

The weight people are actually carrying is less the technology itself than the silent agreement to keep performing as though it has not changed them.

This is what I kept hearing across the day in one form or another, over the second drink and on the walk to the cars. It was an acknowledgment that the speed of this has gotten beyond the capacity of most human nervous systems to process in real time, and that almost nobody is willing to say so in public, because to say so sounds like an admission of not keeping up.

What struck me was how widely distributed the feeling was. It wasn’t the people who had missed the wave who carried this most heavily, but the ones who were ostensibly riding it: founders who had closed rounds this year, operators who had led transformations, engineers who were on paper the ones doing the building. All of them were circling, in different ways, the same private suspicion that the rate of change has started to exceed the rate at which any individual can remain legibly themselves.

I have been thinking about that hour for weeks. Nothing was solved in it, but something was named, and what got named was quietly destabilizing: that most of us are performing a version of ourselves who can keep up, that almost none of us actually can, and that we have entered into a kind of collective agreement not to say so.

Underneath that admission sits another one, harder to make out loud.

We are racing, partly, toward a world in which we hope to finally stop racing.

Some part of all of us is hoping the machines will take enough of this grind off our hands that we can put the phone down and step back into our own lives. And because we have been building identities around that grind for so long, we are not sure, when the moment comes, whether we will know how to let go of it.

I watched all of this surface in small ways during the day, in how people spoke when the cameras were off, in the tired laugh after a candid sentence, in how long the silences lasted, in how grateful people seemed afterwards to have heard someone else say the thing first. The conversations did not end with answers but with the specific kind of relief that comes from being met.

Most of us are carrying some version of this: a pilot that has run for six months with nobody willing to kill it and nobody willing to scale it, an AI strategy slide rearranged more times than we would like to admit, a question from a board member or an employee that we answered more confidently than we actually felt. Underneath all of it sits the same private suspicion, that we are running faster than we can actually live, and that we are not sure what we would do with ourselves if the running ever stopped.

I do not have a resolution to offer, only the memory of that day, of a room full of people who finally spoke plainly, and a reminder that the version of this conversation that is actually worth having is almost never the one on stage. Whether any of us will know how to let go, when the machines finally take enough off our plates for us to try is a question I suspect is living quietly in more of us than will say so. I think many of us are looking for more rooms like that one, even if we cannot yet name it.

Quieter rooms,

Yen