I got my first computer in the mid-1980s. A 286 with a 12 MHz processor. I was a kid in Silicon Valley, and that machine opened a window I never closed. I ran BBS software on it. I learned what it meant for a computer to connect me to another human being on the other side of a phone line. That was digital's original promise — not that the machines would replace us, but that they would close the distance between us.
Somewhere along the way, we forgot that. The industry built extraordinary technology and built it on a foundation of transactions, metrics, and optimization for things that could be measured — while quietly sacrificing the things that couldn't. The click-through rate mattered more than the relationship. The conversion funnel mattered more than the customer. The scale mattered more than the soul.
And then AI showed up, and a lot of people looked at it and asked the wrong question. They asked: "How much of the human can we automate away?" I think that's exactly backwards. The right question — the only question that matters in 2026 — is: with AI doing the rote work, how do we use the freed-up human capacity to make every interaction feel more personal than ever before?
The thesis, stated plainly
Here's what I actually believe, stated as simply as I can:
1. Humans want to be known. Every single customer interaction, at its core, is a human being deciding whether a company sees them as a person or as a number. This is true in B2C and it's true in B2B — the buyer is a human, no matter what's printed on their business card.
2. Digital tools removed friction but also removed connection. The same automation that made it easier to buy anything from anyone also made it feel like nobody was paying attention. You can now purchase from 10,000 brands you'll never remember, because none of them remembered you first.
3. AI is not the enemy of humanity. Lazy implementation is. The default use of AI is to compress costs and scale the least-human version of your interaction. The winning use of AI is the opposite — to compress the cost of the boring parts so humans can focus on the parts that actually matter, which are the ones AI can't do.
4. In a world where everything can be automated, the human touches become the moat. The brands that will own the next decade are the ones that use AI to do the rote work and then spend the freed-up human capacity on the moments that can only land when a real person is involved.
That's the thesis. Everything I've built, every company I've led, every piece of advice I give to operators — it all comes back to that.
Where I've seen this play out (and where I've seen it fail)
In my operator days running a roll-up of digital services firms, the industry was racing to standardize, templatize, and commoditize. The companies we acquired that thrived after the deal were the ones that kept their "small shop feeling" — the account manager who knew your kid's name, the designer who pushed back on your brief because they cared. The ones that didn't survive integration were the ones where we let the standardization process strip out the human parts in the name of efficiency. I watched it happen in real time, more than once, and I learned the lesson the expensive way.
At the beauty marketplace platform I'm building now, the thesis is the architecture. On paper, it looks like a marketplace. In reality, it's a trust network. Connecting 632 Korean brands to 330 Latin American buyers is not a listing problem — it's a relationship problem. The tech runs the rails. The humans build the trust. If I tried to automate the relationships away, I'd have a feature set with no soul, and the users would leave within a quarter. The value lives in the human part. The AI lives in the rails.
AI is going to automate 80% of what used to be your competitive advantage. The remaining 20% — the human part — is where the entire margin will live. If you haven't already started treating that 20% like it's the product, you're behind.
The AI-era paradox
Here's the part people keep getting wrong. You'd think that in a world where AI can write, design, code, sell, and support at scale, the human element would become less valuable. The opposite is true. In an AI-saturated world, anything that obviously came from a human — a real one, with a real voice, who really paid attention — becomes disproportionately valuable precisely because it's rare.
The economics of attention invert. When machine-generated content is infinite, human attention becomes a gift. When machine-generated writing is everywhere, human-crafted writing stands out. When everyone's support chat is a bot, a human voice on the phone feels like luxury. The brands that understand this will spend the next decade winning on the basis of things that don't show up in their AI budget line.
This is not a rejection of AI. I'm on record as one of the most aggressive AI adopters I know. I run my entire operation through Claude Code. I have scheduled agents running while I sleep. I've written about the Force Multiplier framework elsewhere. I'm all in on the tools. The point is that AI is the rails — and the humans are the product. Get that order right and you build something durable. Get it backward and you build a commodity.
What this looks like in practice
If you want to operationalize "Making Digital More Human" in your own company, here's where I'd start:
1. Audit your interaction surfaces for humanity. Every place your company touches a customer — the website, the email flow, the support channel, the onboarding, the billing, the offboarding. Rate each surface 1-10 on "does this feel like a real person cared?" The low-scoring surfaces are where you're leaking loyalty every day without realizing it.
2. Use AI to remove the friction, not the person. The right application of AI in a customer-facing business is to strip out the administrative tax — the forms, the data entry, the back-and-forth, the coordination overhead — so that your humans can spend their time on the moments that matter. If you're using AI primarily to replace the human moments, you're doing it backward.
3. Invest in the moments that can't be faked. The handwritten note. The phone call when the deal is in trouble. The in-person visit. The follow-up that references a specific thing from the last meeting. These scale poorly, which is exactly why they matter in a world where everything else scales well.
4. Measure what matters. Most companies measure the easy things — conversion, churn, engagement. They don't measure the hard things — whether a customer felt seen, whether an employee felt heard, whether a partner felt respected. The hard things are the leading indicators. The easy things are the trailing ones. If you only measure the easy stuff, you'll be surprised when the trajectory shifts on you with no warning.
The challenge
Where has your company let the digital replace the human — and what's it costing you?
This is the audit I'd run if I were walking into your business tomorrow. Not "where can we add more AI," which is the question everyone is asking. The better question: "Where have we let the efficiency of digital strip out a human touch that was actually the thing our customers were paying for?" That's the leak. That's the place to start.
Making Digital More Human is not a nostalgic argument. It's a forward-looking bet on the shape of the next decade. The technology is going to keep getting more capable. The moat is going to keep shifting toward the things the technology can't do. The operators who build their companies around that truth will look, in retrospect, like they saw something the rest of the industry missed. They didn't miss anything. They just remembered what digital was for in the first place — closing distance between humans, not replacing them.
I started this bet in the mid-1980s on a 286 machine in a Korean American kid's bedroom in Silicon Valley. I'm still running it. I think I'm more right about it now than I've ever been, and the back 9 is going to prove it.
Do what you say, say what you do. Make the digital more human. The rest takes care of itself.