What's Left for Humans to Do?

On The Good, The Bad & The Possible

I’m writing to you this morning from the green-gray shoreline of Salinas, California, the place Steinbeck grew up. One of my best friends lives here and I’ve come to visit him, revisit Travels with Charley, and watch salt water remind me once again that I' am not the center of the earth’s gravity.

And yet, I’m taking a break from taking a break to write to you. Why?

Well, first, because I have a new podcast episode to share.

But second, because last night that friend and I talked about pretty much everything we disagree about: religion, politics, and the fundamental way we see the world. It was passionate. Awkward at times. Humbling at others.

It’s funny. I came up here to escape a question that’s just won’t let me go…one that’s been lingering in the back of my mind this entire year like a ghost. Or perhaps a fly. Something haunting and uninvited on my burger.

The question:

Now that we’ve all gotten comfortable with robots running things, what’s left for humans to do?

If you’re leading anything right now (a brand, a team, a project, future generations) I’m guessing you’ve asked some version of this lately. AI can write a strategy, draft a campaign, generate a framework, even mimic Steinbeck. So what’s the work of humans?

This morning my friend poured me a cup of coffee.

“You okay?” he asked.

“Yeah. You?”

“Totally. Love you.”

“Love you too, buddy.”

And then we kept talking. Because the answer to the question, “What’s left for humans in the room?” is equally simple and increasingly difficult…

The short answer: stay in the room.

The longer answer, it turns out, is the subject of this week’s episode of Reculture.

The thing that’s becoming increasingly clear is that it doesn’t matter whether we’re talking about business, society, friends, or family—the hard part is no longer coming up with clear language. Language has now been made cheap. Anyone can generate it. And it only seems to go so far in persuading people.

The hard part is keeping people moving together toward what matters most when they interpret the same language differently.

I’ve spent months researching, analyzing, and trying to make sense of how we work and communicate in this new age of synthetic production. This podcast episode is my attempt to distill everything—the good, the bad, and the possible—into about fifteen minutes. If you listen to anything I’ve put our this year, I humbly hope it’s this.

Listen, watch, or read the transcript below.

---

This One Time in Tokyo

Okay, so maybe you are like me and you’re leading people and you’re making things, and you’re trying to do it in a meaningful way so the story sticks and it scales and it grows beyond you. And maybe you’ve been asking this question that I’ve been asking myself lately: Is this kind of work, work focused on brand and meaning and alignment, designed for a workforce that might not exist in the same way ten years from now?

If companies are automating more and more, if AI can make podcasts and write campaigns and summarize strategy and generate frameworks, what is left for humans to actually do? What is left for me to do? I think we’ve all been asking ourselves a version of this question lately. So buckle up. We are going to attempt to answer it. Actually, it’s going to go deep, so stick with me. But first, a story to start us off.

A few years ago, I was in Tokyo working with a company on brand strategy, and we had an interpreter. The stuff we were doing was heavy, important stuff. Everyone needed to be on the same page for this. Everybody needed to understand the stakes involved. What was weird was that language I thought was very simple in English would take the interpreter an extremely long time to translate to the team. There would be lots of dialogue back and forth, and sometimes I would just kind of sit back and wonder what they were discussing. I don’t know, maybe they were making fun of me or something.

But on the other hand, some of these concepts I had been working on, some of these concepts I had been struggling to communicate clearly to companies in the U.S., stuff that felt really complicated to me, I would be super verbose and the translator would try to translate. Then the team would come back really quickly and say, “Yeah, we have a word for that.” And it would be this one-syllable word in Japanese that perfectly described the thing I was trying to say.

A simple but really important lesson is going to shape the rest of this podcast: language and meaning are not the same thing. Language and what language means to people are not the same thing. Language is becoming super easy to scale. No doubt about that. We have these large language models. Meaning, I think, is actually becoming harder.

What AI Makes Easier for Leaders and Teams

So let’s talk a little bit about the stuff that is legitimately getting easier first. AI is extremely good at two things: pattern recognition and symbol manipulation. AI can predict patterns in language and then recombine words based on probability. It’s really good at drafting, summarizing, and basically formatting at scale.

If you give it five different versions of a company strategy, it can do pretty amazing things, like detect common themes and generate four clean narrative options. Then it can even refine those options based on your feedback. So here’s the brass tacks: it can produce something that sounds pretty coherent. In many cases, it kind of sounds better than what a team of humans can do on its own.

To me, that is real, legitimate value that I don’t think anyone should ignore at this point. Because for CMOs, that means a new level of efficiency. For founders, that means speed. For lean teams like mine, it’s access to things that were harder to access before. And yes, AI can absolutely reduce the need for certain kinds of external support: drafting contracts, generating content calendars, polishing language, summarizing research. Those layers are all squeezing. Pretending otherwise would be naive at this point.

Organizations Are Not Language Systems

Look, here’s the part I think matters. Even if AI can produce a super clear story, and again, even if leadership says, “Yeah, no, that all sounds right,” it doesn’t mean all the people that organization needs to survive will interpret it the same way.

Organizations are not language systems. Organizations are coordination systems under uncertainty. Coordination doesn’t depend on just pattern recognition. It doesn’t just depend on symbol manipulation. It depends on shared interpretation.

AI can align the words. It cannot align incentives. Meaning, it can’t resolve the fact that different people are rewarded, evaluated, and promoted based on different definitions of success. AI can surface those patterns, but it can’t negotiate status. It doesn’t navigate power dynamics or ego or fear or hierarchy. It doesn’t navigate the subtle human calculations happening in the room about who wins, who loses, and who gets blamed if things fail.

Why Clarity Does Not Create Commitment

AI can generate clarity. It can’t generate commitment.

And here’s why I think this is particularly important. If you lead a team, you know this. The real fracture point in most organizations is not in the drafting stage. It’s in the interpretation stage. It’s that moment after five super smart people walk out of the same meeting with five slightly different understandings of what they just heard.

That gap is not a language problem. It’s a meaning problem.

This has actually frustrated me over the years because it is ironic to me that meaning is kind of an ambiguous term. So I’m going to attempt to explain what I mean by meaning. In short, meaning is what counts as the right thing.

To be fair, AI can help clarify that. It can ask you better follow-up questions. It can prompt you to be more specific. It can draft more detailed definitions. It can honestly even expose contradictions in how you might be thinking about or describing your values. For all the talk about these LLMs being sycophantic, I have found that if you push it to push back, it will push back. But again, clarification is not the same thing as commitment.

The Moment Meaning Becomes Culture

For example, if a company says it values innovation, what actually counts as innovative? Is it speed? Is it risk? Is it disruption? Is it efficiency? AI can help you describe that more precisely, but it opens up another question, and I think it’s the more consequential one in real life: What will we choose when innovation costs us something?

If a leader says, “Prioritize people,” okay, what does it mean to prioritize people? Does it mean flexibility, accountability, performance, empathy? AI can generate a clear policy. It can structure a better statement. But culture is revealed when revenue drops, when pressure rises, when there are trade-offs in play. That’s the moment where meaning stops being language and becomes culture.

Where Human Work Becomes Durable

By the way, obviously there are lots of problems and challenges with AI and those who own these models. I’m not trying to oversimplify anything here. I’m just trying to be helpful and realistic and a little bit hopeful in an era that feels kind of disorienting. Also, giant asterisk: none of us know where this is going completely. I don’t know where this is going completely.

But here is where I see the opportunity for human work that feels durable. It’s not replacing AI. It’s not even really competing with AI necessarily. It’s operating at the moment where meaning becomes consequential.

The work is no longer, “Can we draft a better story?” The work is, “Do we all mean the same thing by this story before we ship it, before we scale it, before we build an entire campaign around it?” Because once a story leaves the room, it stops becoming language and starts becoming something people have to engage with. It becomes something people actually have to live inside.

If meaning fractures at that point, then execution fractures, culture drifts, and trust erodes. Meanwhile, AI is just kind of sitting there. AI doesn’t feel that fracture. Humans feel the fracture. So I think that’s where humans should put their energy, especially those who consider themselves brand people or story people or culture people.

What Happens When Organizations Become Software

Just to make sure we’re not being all pie in the sky about it, I want to call out something I’ve been worried about. Something I feel is both a risk and potentially an opportunity. The obvious existential question we’re all sort of dealing with is: What if an organization just automates every human out of it to maximize profit? What if that’s where we’re ultimately headed? What if we’re just all those kind of doughy humans riding around in scooters like in the Pixar movie WALL-E?

So here’s my take. If an organization could truly reduce itself to just pure automation, it would actually stop being an organization and simply be software. And here’s the thing: I for sure think that’s going to be the goal. Many actually already have done this. I think it would be naive to think that we’re not going to continue to see that trend rise, that people aren’t going to try to get humans completely out of the loop.

But I could be wrong. I don’t think we’re going to suddenly stop needing organizations altogether. Why? Because remember, organizations are groups of people who can coordinate systems under things like uncertainty and volatility and change.

Culture as Coordination Under Uncertainty

I don’t see those things going away anytime soon. The moment people are part of the equation, the difficulty shifts from execution, which is the easy part. That’s AI. That’s software. It shifts to coordination. And coordination is meaning dependent.

Coordination is the thing that raises its hand and asks, “What actually counts as the right thing around here?” And here’s where this is going. I think the answer to that question actually has a name. It has something we can point to. It has something we already have language for: culture.

Culture is not perks. It’s not posters. It’s not the values page on a website. Culture is the collective, shared answer to the question, “What does this mean?” What counts as the right thing around here? What counts as success? What counts as risk? What counts as loyalty? What counts as failure?

And those answers aren’t generated by better sentences. They materialize through shared meaning. Shared meaning requires tension. It requires clarification. It requires conversation in real life. And more often than not, at least in my experience, it requires a little courage.

Why Output Is No Longer the Deepest Work

So here’s the shift I think is actually happening, plainly said. For years, we thought the work was output. We thought it was more content, more clarity, more consistency. But the deeper work is durability.

Does the story hold together once people start acting on it in real life? Does it survive well? Does it walk on its own two feet when it’s exposed to human contact?

That’s post-AI. That’s about contact and connection. And connection is social. AI is only going to get better. It’s going to draft faster. It’s going to summarize better. It’s going to sharpen and clarify sentences.

The Future of Brand, Strategy, and Storytelling

And if I’m a betting man, it’s going to notice a lot more than we expect it to. I think it’s probably going to do some very human-like things. Maybe it will detect hesitation or translate emotion. I don’t think any of us really know where this is going. But here’s what seems to be true so far. Even if AI can help articulate values, even if it can help refine trade-offs, even if it can help pressure-test strategy, someone still has to decide.

Someone has to bear the cost. Someone, at the end of the day, has to live with consequences.

So the future of work we call strategic or creative, things like storytelling and branding, is probably not going to be about producing more language. That’s still helpful, but it’s been downgraded. It’s probably going to be more about stewarding a moment, the moment where language becomes reality.

It’s going to shift from producing a bunch of messages to stabilizing meaning. It’s going to transfer from generating a bunch of ideas to helping those ideas survive contact with human life.

Why Social Skills Become Strategic

This is where I actually think things get really interesting. This is where I get really hopeful because I think we are approaching a kind of paradox moment. For years, we’ve been worried that technology is making us less social and more isolated. And I think that is 100% true.

But if the future of human work depends on stabilizing meaning under pressure, on clarifying it, that’s going to take people with some pretty strong social skills.

It used to be, if you wanted your kids to get ahead in life and in their career, what would you do? You would teach them to code. Now I think it looks more like teaching them to decode. To decode humans. To know how to solve conflict. How to understand different perspectives. How to look for nuance.

Those skills become strategic. They become rare. They become scarce because you cannot coordinate meaning without genuine, engaged, empathetic conversation. You cannot build commitment without trust, something that gets learned and earned and worked out in real life. You can’t navigate trade-offs without some kind of relationship.

From Content Production to Social Infrastructure

I think the future will still belong to those who can leverage their creativity to aim people somewhere. Absolutely. But I think it’s starting to look a lot less like content production and more like social infrastructure. A lot less like social media and more like social work. Not necessarily in the bureaucratic sense of the term, but in the human sense.

To name the tension. To surface the misalignment. To keep people in the room, listening, paying attention when the stakes are high.

I think the felt need from leaders is going to shift from, “I need someone who can help me share my story,” to, “I need someone who can translate intent when it’s getting distorted by all of the generated noise around us.”

What Work Remains After AI Can Say Almost Anything?

So I don’t think the question is, “Will AI replace creative work or strategy work or fill-in-the-blank kind of work?” I think the better question, the better framing, is, “What work remains necessary once you can automate whatever you want to say?”

And the answer, as far as I can see, is this: It’s the work of making sure what we say is what we mean. And that what we mean equals the thing that actually shows up once people start living inside it.

That’s culture. And that’s where the future of work lives.