Fixing the Disambiguation of a Startup Founder
What a 5-LLM audit at Day 8 revealed about how machines decide who exists, and what it takes to cross the line from private individual to recognized entity.
The situation
You don’t exist in machines yet.
A proptech founder I know reached out recently to get my advice on his visibility strategy. He’s in the middle of a soft launch of his platform: onboarding agents, building relationships, and making sure the platform is available in app stores for both Apple and Android users. The platform is solid. It’s a digital real estate marketplace for Guyana, with ambitions to replicate across 51 countries. He taught himself to code at 52. He has traction. And he’s thinking about press, visibility, and speaking opportunities.
Before any of that could work, I told him something he didn’t expect: you don’t exist in machines yet. And that’s going to matter for building trust far more than a one-off appearance on Good Morning America.
The problem wasn’t that he was unknown. The problem was a shared name, and a more visible person already owning it. At least four people show up when you search that name. The most established is a Portland-based AI operator whose entity had more authority signals, more trust, and more history with the machines. So whenever the name was queried, the machines defaulted to him. Search for the proptech founder, get the fractional AI operator instead.
That’s a disambiguation problem. And it had to be solved before anything else.
The intervention
Four moves. Eight days.
The audit · Day 8
Five systems. Five different answers.
I ran the same query across five AI systems in a single session: Who is Darren L Buckner? What came back wasn’t just five different answers. It was five different epistemologies, five different logics for how a machine decides who you are.
“If the internet can confirm you without you, you’ve crossed the line.”
ChatGPT, April 20, 2026 · on what it would take to reclassify a private individual as a public figureThe finding that matters
Perplexity showed us the inside of the machine.
When I asked Perplexity to differentiate confidently between the two Darren Buckners, it didn’t just give me an answer. It named the mechanism it used to arrive at one.
That combination points to exactly one person. No other Darren Buckner has that cluster. It’s like a fingerprint. AI systems don’t resolve entities through prestige. They resolve through corroboration: distinctive signals, co-occurring, across sources they trust. And Perplexity revealing that cluster was an unexpected answer from the LLM I often call the fastest learner.
Eight days of work was enough to build that fingerprint. Not yet enough to propagate it everywhere. That’s the gap we’re still closing.
Key takeaways
What this tells us about how machines decide who exists.
Private is the default. You have to earn the reclassification.
Every person starts as a private individual in the eyes of AI systems, legally and epistemically. The burden of proof runs one way: you have to build enough signal for the machine to feel safe surfacing you.
Different systems are running different logic.
Copilot trusts LinkedIn. Perplexity clusters signals. ChatGPT defaults to privacy frameworks. Gemini resolves to authority, even when the authority is wrong. A single strategy won’t work across all five.
The middle initial is a search key, not just a brand choice.
“Darren Buckner” is a contested namespace. “Darren L Buckner” is not. Eight days in, the search term is clean. The entity isn’t fully assembled yet. That’s the next stage.
Corroboration beats prestige.
A Forbes mention won’t help if it doesn’t include the right identifying signals. What moves the needle is a cluster of co-occurring, distinctive details across sources the machine trusts. Build the fingerprint first.
The gap between findable and known is a propagation problem.
He’s findable today. His knowledge graph isn’t assembled yet. Those are different problems with different timelines. The intervention happened. The indexing catches up on its own schedule.
Are the machines confusing you with someone else?
Find out what AI systems actually say about you, and what it would take to fix it.