Merge Conflict

I googled myself on my phone this morning, which I don’t recommend.

Google’s AI Mode had written me a biography, and to be honest, it was a pretty good one. I graduated from Georgetown in 2012 with a degree in History and Classics. I earned an MFA in Film and Television Production from USC in 2016. I co-wrote and directed a feature that went to Netflix, and I founded two production companies in Albuquerque.

The only problem is that I didn’t do any of that.

Those achievements belong to a filmmaker in Albuquerque whose name is mine plus one letter. The AI knew about the letter, too. It listed his spelling as an alternate credit for me, the way you’d mention that a novelist sometimes writes under a pen name.

To be fair to the machine, my name twin and I don’t make it easy. I grew up in New Mexico. I have film credits of my own, including executive producer on The Lost Husband, which went to #1 on Netflix. So when an algorithm finds two guys with nearly the same name, both tied to New Mexico, both with movies on Netflix, it does what any of us would do at the end of a long day. It decides they’re the same guy. (Georgetown’s class of 2012 would mean I was either in my mid-thirties as a freshman or I’m in my mid-thirties now. I wouldn’t mind being in my mid-thirties now.)

The desktop version of the same search got me mostly right, which somehow made the phone version worse. It meant the facts were out there. The machine just had two piles and couldn’t tell which one was mine.

My own website wasn’t helping. It tells my life story as a software changelog, with modules and patches and reinstalls, and a human reader gets the joke. A machine doesn’t see a joke at all, much less understand it enough to get it. The page never said where I went to college, so when the AI needed a college for me, it borrowed his.

The fix turned out to be simple: write with an AI reader in mind rather than a human one. I added a block of structured data to every page of this site. Visitors never see it, but search engines read it, and it says who I am, where I studied (Middlebury, then an MFA in Creative Writing from the University of Houston), and which profiles on other sites are actually mine. It also includes one sentence I never expected to write: I did not attend Georgetown or USC. I wrote a plain fact sheet for AI crawlers. (Google says it doesn’t read these. I’m keeping it anyway, for the other machines.) I registered the site with Google Search Console and asked Google to take another look. Then I filled out my IMDb page, which until this week listed my credits and nothing else. Not a bio, not a school, not a photo.

The honest status is that I don’t know yet whether any of it worked. Google updates these answers on its own schedule, and that schedule is measured in weeks.

I also don’t know whether my name twin has ever googled himself and found my resume instead. If he did, I imagine he was as surprised to find out that he was a novelist who founded a grammar company as I was to learn that my MFA was in Film and TV.

In the meantime, my IMDb page has a face on it for the first time. It’s mine. I checked. (Twice.)

---

If you have a name twin

Here’s what I did, in the order I’d do it again, with what I know and don’t know about each step.

1. Put the facts where people can see them. The machine borrowed a college for me because my site never named one. Say the basics (where you studied, what you do, where you live) in plain text on your own site. This matters more than I realized: Google’s guidelines say structured data should describe content that’s visible on the page, so the hidden version works best when it matches the visible one.

2. Describe yourself in structured data. It’s a short block of code, in a format Google explains here, that tells search engines who you are using the schema.org Person type. The most useful part is sameAs, a list of your profiles on other sites, which ties them all to one person. There’s also a field called disambiguatingDescription that was made for my problem. I haven’t found anything from Google saying it reads that field, so consider it a bet.

3. Check your work. Paste your URL into the Schema Markup Validator or Google’s Rich Results Test. If something’s broken, they’ll tell you.

4. Ask Google to look again. Add your site to Google Search Console, submit a sitemap, and use URL Inspection to request indexing of your homepage. It won’t make Google change its mind on your schedule, but it makes sure Google knows there’s something new to read.

5. Claim your knowledge panel. If Google shows a box about you on the right side of desktop results, you can claim it and suggest corrections. Mine calls me a “Film producer,” which is technically true and is also the whole problem.

6. Fix the profiles machines trust. Search engines lean on big sites to decide who’s who. For me that meant IMDb, which had my credits and nothing else. For you it might be LinkedIn, an Amazon author page, Goodreads, or a company bio.

7. Optional: write an llms.txt. It’s a plain-text fact sheet for AI tools that lives at yoursite.com/llms.txt. Mine is here. Google says its search features don’t use it, but other AI tools might. It’s short, and it costs nothing.