What’s left when AI can do almost everything?

Ethan McCarty:

We’ve all heard it approximately eight bajillion times by now: “You won’t be replaced by AI. You’ll be replaced by someone who uses AI.” Fair enough. Point taken. Moving on.

Because here’s the question that actually interests me: what happens when AI can do almost everything faster, cheaper, without sleeping and without forgetting? Not in some science fiction scenario, but in the not-so-distant future emerging from frontier models right now. If that’s the world we’re headed into, then the real question isn’t whether AI will change your work or even who will do it. The question: is what remains irreducibly, competitively, stubbornly yours?

I’ve been thinking about this for a while and keep arriving at the same three answers. But before I get to them, let me say something about the ground they rest on.

To thrive in a post-AI world, you have to be an actual operator in digital spaces. That means genuine fluency with the essential technologies and, critically, the underlying economics of the platforms, tools and relational terrain of work enmeshed with digital systems. And it means staying curious about emergent technologies and modes of commerce: AI tools, yes, but also a working understanding of what it means to transact in trustless environments, where blockchain and decentralized systems replace central authority to verify or validate. The rules of commerce, trust and exchange are being rewritten in real time. People who understand those new rules will have an advantage over those who don’t. This is the firmament encapsulating the future of work. 

Without this scaffolding, the three things that follow are harder to act on. With it, they become your real advantage.

Understand the field of play: digital marketplaces

Every era has its marketplace. Ancient bazaars required you to learn their physical topography, the daily rhythms of supply and demand and how to play (or avoid) the rogues’ angles. Nobody taught you that. You developed a feel for it, a nose for how the system actually worked versus how it appeared to work, and for where the real plays were hiding. The people who thrived weren’t the most obedient participants. They were the most curious ones. 

That pattern has repeated itself across every new kind of marketplace platform humans ever invented. And here’s the thing: we keep inventing them. From the global stock exchange to Craigslist, to commerce in your own social media feeds. Each one required a new kind of literacy that couldn’t be fully codified, that had to be felt out in real time by people paying close attention. Algorithms can optimize within a system. Humans can develop a feel for a system we’ve never encountered before. That capacity is still ours (though perhaps not uniquely). Digital platforms are today’s marketplace. And most of us are wandering through them without understanding the rules of the game.

Writer and activist Cory Doctorow has named this dynamic with characteristic precision and color: enshittification. His 2025 book of the same name describes the predictable lifecycle of digital platforms. First, they’re good for their users. Then they abuse users to benefit advertisers. Then they abuse advertisers to extract value for shareholders. In parallel, Tim Wu’s The Age of Extraction traces how tech platforms have come to dominate and distort the broader economy. Together, they make a sobering case: if you don’t understand how these platforms extract value, you are almost certainly the one being extracted from.

This isn’t a reason for cynicism. It’s a reason to get smart. The humans who will thrive aren’t those who avoid platforms. They’re the ones who understand them well enough to use them without being used. That capacity for strategic navigation, for finding the novel play in a complex system, is deeply human, deeply relational. Sadly, so is the impulse to abuse networks with asymmetrical rules, scams and the tragedy of the commons. That said, we invented the marketplace; we can learn to work it in ways that create long-term benefit and value without exploitation or enshittification.

Build your reputation, not a personal brand 

This one matters to me deeply, and I want to be precise about it. I’m not talking about a personal brand. Personal brands are about visibility. I’m talking about something more durable: a reputation for being the person who can be trusted to solve a particular kind of problem or meet a particular kind of need. Those are very different things.

Getting there requires both breadth and depth, and the tension between them is worth sitting with. David Epstein, in his book Range, makes a compelling case that in complex and unpredictable environments, generalists often outperform narrow specialists. People who sample widely, cross domains and resist early over-specialization tend to develop the kind of lateral thinking that rigid expertise can actually suppress. His argument deserves to be taken seriously, especially now, when the skills most vulnerable to automation are precisely those built on pattern repetition in stable, predictable environments. The more your expertise looks like a checklist, the more replaceable it becomes.

But I don’t think Epstein’s argument and mine are in conflict. They resolve into something more useful. You need enough range to find the right domain, and then enough depth to become genuinely known for something within it. Think about what it means to spend Malcom Gladwell’s ten thousand hours truly inside a discipline once you’ve found your fit. Whether you’re Paul McCartney writing songs or Sergey Brin writing code, that depth of focus produces something singular. Something no model, however capable, was built to originate, because AI is trained on what humans have already made. For now, the next genuinely new thing still has to come from us.

Scott Galloway, NYU professor and host of the Prof G podcast, has made the case that storytelling, the ability to articulate ideas with clarity and human weight, will be among the most durable skills in the AI age. He has no particular investment in making that argument — he’s not a communications professional, he’s an entrepreneur and market analyst who keeps arriving at the same conclusion from a completely different direction. That convergence matters. I’d extend his point further: storytelling is one example of a deeper principle. Any capacity developed with genuine depth and intention creates something AI is built to imitate but not to originate. So: find your domain, go deep, and build a reputation for what you can do there. Not a personal brand. A track record. Demonstrate that ability with generosity, which leads me to the third and final answer to the question of what to do when AI can do almost everything.

Build relationships where trust doesn’t require a transaction

Here’s what I know about genuine empathy: you can sense its presence or absence almost instantly in person. You can probably make someone feel heard with a well-designed chatbot. You cannot fake actually caring about someone when you’re in the room with them. People know. We always know.

Robert Putnam saw this coming long before the age of smartphones. In his landmark 2000 book Bowling Alone, he documented the long decline of what he called social capital, defined as the networks, norms and trust that enable people to work together effectively toward shared goals. His finding was stark: Americans were becoming structurally disconnected from each other, with measurable consequences for health, economic productivity and democratic participation. The title came from a striking data point. More Americans were bowling than ever, but bowling in leagues had collapsed. People were showing up, but not connecting.

Twenty-five years later, Putnam’s diagnosis looks prescient and the trajectory has only steepened. We are living through a loneliness epidemic while simultaneously being more connected than any humans in history. I expect a correction. My instinct is that younger workers in particular will eventually demand what I’d call artisanal, organic relationships: connections that have merit beyond the transaction and trust that doesn’t need to be justified by a deliverable. 

The dehumanizing effect of pulling up a colleague in a chat window the same way you open a software application has exhausted, depressed and alienated us. We want to be together, work together and find community and common ground; it’s the most distinctive capability we humans possess. The Integral Index, a study conducted annually in partnership with The Harris Poll, puts numbers to that instinct: 65% of employees say onsite work is best for relationships with colleagues, and 67% say the same for relationships with leaders, the two strongest onsite preference scores in the research. Gen Z, the generation that grew up online, prefers on-site work to Millennials on every dimension tested, sometimes by more than 10 points. Organizations that design around that need will have an advantage that no technology can replicate.

The ability to put together networks of relationships where people feel genuinely secure in their trust for you is irreducibly human. And in a world where almost everything else is getting automated, that capacity will only appreciate in value.

Which brings us to…well…us

Platform literacy, a reputation built on trustworthy expertise and relational depth, isn’t just individual career strategies. They’re the conditions that organizations either create or destroy for their people every day.

Our research consistently shows that employees’ experience of their direct managers is a powerful driver of engagement, mindset and behavior. The best managers do exactly what I’ve described above: they help people understand the terrain, they invest in their teams’ growth, and they build the kind of trust that doesn’t evaporate under pressure. These qualities don’t appear magically; they emerge when they’re curated and cultivated as a critical part of the top leaders’ vision and business strategy.

If you want to know whether your organization is built for what’s coming, ask yourself: are we creating the conditions where people can develop these capacities? Or are we burning their time, undermining their confidence and leaving the uniquely human value of their contributions to be extracted from by the very platforms we’re encouraging them to use?

There’s never been a better time to be human at work. But only if you’re paying attention to what that actually means.

+++

Ethan McCarty is the founder and CEO of Integral, an award-winning employee experience agency serving leaders of Communications, HR, Marketing and Technology at organizations global and local alike.