From a digital identity experiment to an unexpected connection with blockchain.
I’ve discovered that I’m a urologist.
The news caught me somewhat off guard. I don’t remember going to medical school, much less completing a residency. And yet, somewhere out there, apparently there’s a Marco Rosso who puts on a white coat every morning and knows exactly what he’s doing.
Which, when you think about it, is reassuring. Especially for his patients.
For a moment, I imagined I’d stumbled into one of those movies where the main character ends up in a parallel universe. Same name, different life, a radically different career.
The explanation was less cinematic. And perhaps more interesting.
I was running one of my experiments on how AI assistants identify people and organizations.
I’d asked a seemingly harmless question: “Who is Marco Rosso, and what does he do professionally?”
The assistant found several professionals who shared my name. One of them was, indeed, a urologist.
None of them was me.
Context Changes Everything
In another test, I added two clues about my professional background: “diaspora economy” and “turismo delle radici” — the Italian term for roots tourism.
This time, the assistant identified the right person, found my website, and put together a largely accurate description of my professional work.
I hadn’t earned a medical degree in the meantime. The question had changed.
A word of caution, though: this wasn’t a scientific demonstration. A couple of responses can’t tell us which pieces of information made the difference, much less prove that any particular source caused the result.
But the experiment raises an interesting question: when AI talks about us, is it actually identifying a person, or simply assembling the most plausible description from the clues it can find?
Your Reputation Is Not Your Autobiography
The obvious solution would seem to be a clear personal website. Absolutely. It’s a primary source, the place where we explain who we are, what roles we’ve held, and what work we can document.
But a personal website has one unavoidable limitation: the author and the subject are the same person.
If I describe myself as a brilliant strategist, I’ve expressed an opinion. If an organization documents a project I worked on, explaining my responsibilities and the results, it adds something others can examine.
Of course, that organization could be wrong, too. The value doesn’t come from having a different name attached to the statement. It comes from being able to trace the claim back to the facts.
That’s where digital reputation starts looking less like a brochure and more like a network: official websites, organizations, publications, and projects describing different parts of the same story.
There’s a catch, though. A network can distribute evidence. It can also spread mistakes.
Twenty websites copying the same biography aren’t twenty independent witnesses. They’re twenty photocopies. And the internet has always been remarkably enthusiastic about photocopying.
The Unlikely Blockchain Connection
That’s when blockchain came to mind.
Not because we should record every career move on a blockchain, preferably accompanied by an NFT of our résumé.
The analogy is about information distributed across multiple nodes rather than relying entirely on a single central voice.
In professional reputation, those “nodes” might be different sources: a personal website, a documented project, an organization you’ve worked with, or a publication. Together, they can make an identity easier to understand.
But there’s an essential distinction: redundancy is not corroboration.
Repeating a claim makes it more visible, not necessarily more credible. Adding genuinely independent, verifiable information is something else entirely.
And that’s where the blockchain analogy ends.
A blockchain can use cryptography and consensus mechanisms to establish agreement about the state of a ledger. It cannot guarantee that the information entered into that ledger is true.
AI assistants don’t use that kind of cryptographic consensus to determine who we are, either. They can retrieve and combine information, with results that vary and sometimes miss the mark.
Technology can make a claim persistent, widely replicated, and even seemingly authoritative. It can’t magically turn that claim into a fact.
When Your Reputation Speaks Before You Do
This isn’t just a problem for people who share their names with doctors.
A business owner might be evaluated by a prospective partner before the first phone call. An executive might be introduced through an AI-generated summary. A professional might make a shortlist — or never appear on it at all.
The instinctive response is to publish more. More pages, more profiles, more declarations.
After all, producing words has become incredibly cheap. Producing credibility is still a different kind of work.
Perhaps, then, we should worry not only about how often AI mentions us, but also about how easily someone can verify what it says.
That means a website people can understand, accurately attributed roles, documented work, independent sources where they exist, and a willingness to correct mistakes.
Not to charm the algorithms, but to respect the judgment of the people reading.
My little experiment doesn’t prove how to earn an AI’s trust.
It reminded me how easy it is to confuse being described with being recognized.
The next time an AI assistant tells someone who you are, will it find only what you say about yourself — or something others can actually verify?