This month, I left my role as a senior agent at Writers House to launch an advisory practice for thought leaders. I’ll continue to represent a select list of nonfiction clients at another agency, but I’m closed to submissions.
To explain my reasoning in typically circuitous manner: This summer, I became obsessed with intheweights.com. It’s a website that does one thing: you type in a name, and it tells you how strongly different AI models recognize that person and what they think they know about them. (If we’ve met IRL, I’ve probably told you to look yourself up. It’s been my favorite topic of conversation.)
Since discovering the site, I’ve spent a decent amount of time working on my score and sharpening up my (agentic) image. Call it Radical Narcissism: even the computers’ opinions matter. For example, I might ask Claude for a list of the top five people at doing something I’m good at doing. Then, I’ll ask why I’m not on that list, too. Claude cheerfully lists my perceived failings, and I try to address them. Rinse and repeat.
People call AIs sycophantic, but they can be downright mean with the right prompts. It’s . . . bracing.
Back to the career pivot. As a cultural gatekeeper—editor, agent, etc.—I typically spend half my time sizing people up based on limited information. I need to know: Is this person credible? Do others in their field take them seriously? What’s their deal? (Then, I spend the other half helping my authors, clients, and collaborators look more credible to others. Sort of a “to catch a thief” deal.)
Now, the Machines are asking the same questions I do. That means I can compare my gut impression with a sort-of objective yardstick. Playing with intheweights also gives me a window into how AIs evaluate us, which matters because they’re already replacing Google. When people vet someone’s bona fides now, they ask ChatGPT more often than not.
If you run my name at intheweights.com today, you’ll find that the models classify me as a “book editor and publishing professional,” top 15%. OK, but not great. The picture of me is still askew. In fact, one AI credited me with writing a book called The Art of Eating Through the Zombie Apocalypse. I think I’d remember working on that one.
Cut the LLM sleuths some slack. I get to talk to people before passing judgment on their credibility and expertise. The web offers a very limited view by itself.
As I’ve written before, the lack of a robust online presence has long been a power move among the cultural elite. Social media–savvy writers like Stephen King and Margaret Atwood exist, but many behind-the-scenes players—top agents, media CEOs, etc.—barely have a LinkedIn page. Mapping the real lines of power from a microchip must be incredibly difficult. (I’m sure the agents will start calling people to vet them soon enough.)
I’m not sure how much longer staying out of the fray will fly. GEO (Generative Engine Optimization) is the new SEO, and as AI seeps into every inch of how we make our work (and discover the work of others), I suspect a lackluster digital presence will undermine our professional and creative efforts in unpredictable, often invisible ways.
No matter how much you’ve accomplished, it’s time to start looking in the agentic mirror. A lousy online presence—whether that’s thin or negative—has always been a spinach-in-the-teeth problem. Nobody’s going to tell you they’re passing on your book because of what Google or, increasingly, Claude says about you. The invite to that elite conference or mastermind just isn’t going to appear in your inbox. How do you diagnose the absence of an opportunity?
Now, the gatekeepers checking you out aren’t just Googling your name or business. They’re asking Gemini or Grok (well, not Grok) a specific, nuanced question. I’ve found that the results differ drastically depending on how you phrase your question. “Who are the top five consultants on business development?” versus “I need some biz dev support—who should I reach out to?” How do you anticipate every possible angle of agentic inquiry?
Yesterday’s authors had it made. Most editors and agents are clumsy in their online investigations. They glance at social media, skim a page or two of search results, and largely trust what’s in the marketing and promotion section of the proposal. But AIs?
Meanwhile, today’s aspiring thought leader has a bunch of social media accounts, most of them defunct and out of date; bios from jobs they left years ago that nobody ever scrubbed from the company website; and a random article about a party they threw back in 1998 in the college newspaper. It’s a mess. The skimpier the results, the easier it is for the LLM to hallucinate an entirely false impression out of whole cloth.
Run the test yourself. Let me know what you find, and whether it’s accurate. I’m super-curious about this stuff.
For extra credit, open a private browser window so your history doesn’t tilt the results, and ask an AI, “Who is [your name]?” Use a temporary chat or the LLM, knowing that you’re asking about yourself, will alter the results to flatter you. Write down everything that comes back wrong, missing, or out of date, and then ask the model (logged in this) how to get the spinach out of your teeth. Some of the basic remedies include:
One canonical homepage on a domain with your name.
The same bio language used everywhere.
Links pointing back to your site from every essay, talk, and book—connect the dots, people!
A Wikidata entry that links to relevant sources (yes, that’s kosher; Wikidata isn’t Wikipedia).
All this impression management may feel a little vain. Do it anyway. People verify your record. If you want them to take your work seriously, give them good reason to. Nobody’s going to weigh your ideas on their merit until they’ve confirmed you’re worth the effort.
To bring it full circle: the new practice I’ve founded will help people with real-world experience, credibility, and expertise find a big idea they can own, build a body of work around it, bring that work to the right audiences, and shape a platform that genuinely supports the effort. This isn’t as easy as building a homepage—doing this right takes months, even years—but it’s by far the most effective way to establish your credibility and increase your impact in an era when the people who matter are delegating their judgment to AI.
p.s. If you think you might be a fit for the new consultancy, email me at david@moldawer.com.