For most of the last two years, the story about AI has been about acceleration. What used to take a week now takes an hour. What used to require a team now requires a prompt. What used to cost money now costs pennies.
That story is true. It’s also incomplete.
The part that gets less attention is what happens on the other side of that collapse. When the cost of producing something falls to near zero, the ability to produce that thing stops being rare. Something else becomes rare instead. And in a world where anyone can write anything, generate anything, personalize anything, the thing that becomes scarce is trustability (being worthy of someone’s trust).
Over the last few months, thanks to Founder Real Talk, we’ve talked to founders across completely different industries who’ve independently landed on the same conclusion. AI matchmaking. AI-powered websites. Sales enablement. Agency work. AI safety infrastructure. None of them talk to each other. They’re solving different problems for different customers. And they’ve all arrived at the same thesis.
Trust is the new moat.
Why this is happening now
There’s a version of the “AI is coming for everything” argument that treats every industry as headed toward the same commodified endpoint. Content is commodified. Code is commodified. Design is commodified. Sales outreach is commodified. Everything gets cheaper, faster, and less differentiated, forever.
That framing is missing something important. When a category commodifies, buyers don’t stop caring. They start caring about something else.
You’ve probably heard people in Silicon Valley say, “Taste Is the New Moat.” Their argument: as generative AI collapses the cost of production, the real value comes in curation. Editorial judgment. Point of view. Knowing what’s worth making rather than what can be made.
We think trust is the second half of that same shift. Taste is what someone brings to the work. Trust is what other people extend to them. Both are human. Both are hard to fake. Both become disproportionately valuable in a world where output is free.
Here’s the mechanical version. AI has done two things at once. It has made it much easier to produce content that looks credible. And it has made it much harder to tell whether the credible-looking thing is actually credible. That gap, between what looks trustworthy and what is trustworthy, is where value is now concentrating.
Ben Gusberg at CAPE puts it in the sharpest form. Cold email conversion is collapsing because everyone has the same AI tools. In his words, “everyone is sending 10,000 emails a day that are perfectly personalized to you, and the conversion for those messages is going way down.” His numbers on the alternative are striking. Warm intros convert 34 times higher than cold email.
His framing: “In a world with all noise, how do you find signal? Ironically, we think that signal comes from humans.”
That’s the shape of the argument in one line. As synthetic communication floods the channel, the signal buyers actually respond to is human.
Check out Ben’s full episode on our YouTube channel here
What this means
The founders we talk to who are winning right now are betting on a specific set of things that AI can’t easily replicate.
Real relationships. The kind you build over years, with people who’ve watched you handle the hard version of something and stuck around. This is why in-person is resurgent. In-person can’t be automated or faked. Presence is a costly signal in a way that a cold email will never be, no matter how well-personalized.
Genuine thoughtful voice. Not a brand voice generated by a prompt. A perspective that comes from an actual human who’s spent enough time in a specific problem to have earned an opinion about it. The blog post that took someone twenty years of hard experience to write reads different than the one that took a language model twenty seconds, even when the surface content is comparable.
Reputation over time. In a world where anyone can spin up a plausible-looking company overnight, the founders who’ve been building the same thing for five years, showing up in public, taking responsibility for outcomes, have an advantage that new entrants can’t buy. Longevity is a form of trust that money can’t manufacture.
Safety and audit-ability. The regulatory version of trust. As AI gets deployed in high-stakes domains, the ability to prove your system won’t fail in the ways that matter becomes a hard requirement. The EU is already pushing regulation into force that requires companies deploying AI to demonstrate compliance. Trust is being written into law, not just competitive strategy.
All of these have one property in common. They can’t be produced at zero marginal cost.
What this looks like practically
If you’re a founder trying to decide what to build, or what to invest your reputation in, the takeaway is fairly specific.
Prioritize the parts of your business that can’t be commoditized. The customer relationships. The domain expertise. The judgment that tells you what to bother building and whether what you’ve built actually serves the customers you want. These are the parts that will still be worth something in five years, whatever else changes.
Be skeptical of building anything that depends on a temporary advantage in AI-generated volume. Cold outreach at scale, content flooding, spray-and-pray outbound. These tactics are already declining and will decline further. If your business model requires them to work, your business model is on borrowed time.
Invest in in-person more than the current defaults would suggest. Events, dinners, actual meetings. The economics have quietly shifted. In a world where the cost of digital contact is near zero, the value of physical presence is going up, not down. A dinner with ten of the right people is now doing what a thousand LinkedIn connections used to do.
Build your public presence on the assumption that trust will be the scarce resource for the rest of your career. Not audience. Not reach. Trust. Which means being right when you make a public call, admitting when you’re wrong, and being consistent about what you stand for. That’s a five-year game, and the people who started playing it three years ago are already ahead.
The reframe
The narrative about AI has spent two years focused on what it can do. The interesting question, and the one the founders we talk to are quietly answering, is what AI can’t do.
It can produce output. It can’t produce trust.
It can generate credible-looking content. It can’t give that content the weight of a human being who poured their heart into it.
It can automate the cheap parts of a relationship. It can’t automate the parts that make a relationship worth having.
The founders betting on the human side of that split are the ones we’d bet on. Not because AI is going away. Because the more powerful AI gets, the more valuable the thing it can’t do becomes.
Trust is the new moat. It always has been, actually. The change is that everything else just got a lot easier to imitate.
