Searching your own name used to mean checking what Google could find. In the Weights suggests a different question for 2026: what do AI models already seem to know about you, without using web search?
A vanity search built for the AI era
In the Weights was created by Thomas Dimson and Joey Flynn after they left OpenAI, where they had joined through the acquisition of their design startup Global Illumination. The site is built around a simple but provocative idea: public visibility is no longer only about search results. It may also be about whether large AI models can recall a person from their internal training.
The name points to the “weights” inside AI systems. In the source article, those weights are described as the numerical parameters that shape a model’s training and output. The site says it measures how well “a model is able to recall someone without using tools like web search.”
That makes In the Weights feel familiar and strange at the same time. It has the personal pull of a vanity search, but the results come from chatbots instead of a search engine. The site’s own language turns that into a status signal: “Being in the weights means your existence was deemed important in the process of creating superhuman artificial intelligence,” it says.
How the score is assembled
According to the source article, In the Weights queries different models with a prompt similar to: “Who is <name>? Give up to 10 results, each with a short description and confidence.” The models mentioned include Grok, Gemini, multiple versions of GPT, Claude, and Llama, along with lesser known models.
The site then groups similar descriptions and assigns a strength score. That score makes the result easy to compare, which is part of the appeal. A name does not just appear or fail to appear. It gets ranked.
The article gives one example: Anthony Ha received a strength score of 641, placing him in the top 6% of names. Several TechCrunch colleagues scored higher. During the writing of the source article, the leaderboard was shifting, with “Home Alone” star Macaulay Culkin in the top slot with a strength score of 988, close to opera singer Luciano Pavarotti.
In the Weights also shows which models returned which answers for a given name. That detail matters because the same name can produce different responses depending on the model. It can also expose questionable answers. The source article notes that GPT-5.4 Mini said Anthony Ha is an “ambiguous name form that could refer to multiple people with the initials A.H.A.”
Why people are checking themselves
Dimson told TechCrunch by email that he and Flynn wanted to “get the creative juices flowing again” after leaving OpenAI. He also said he was thinking about how “Google vanity searches are the wrong objective in 2026 as more traffic moves to LLMs” and about “so many lives are encoded somehow in a bunch of floating point numbers inside the AI brain.”
That framing helps explain why the site has struck a nerve. In the Weights turns an abstract anxiety into a visible result. If chatbots are becoming a major way people learn about other people, then being remembered by those systems starts to feel meaningful, even if the meaning is hard to define.
Dimson said the direction of the project was “sealed” by a tongue-in-cheek blog post riffing on AI weights and Terry Bisson’s classic short story “They’re Made Out of Meat.” He also told TechCrunch: “Reception has been insane so far, we thought this would be a mild curiosity but it seems like it has struck a nerve of wanting to see if you live forever in the super intelligence (the comparison factor doesn’t hurt either!)”
The comparison factor is important. A raw chatbot answer can be interesting, but a leaderboard makes it social. A strength score gives people something to react to, argue with, and measure against other names.
The limits of being remembered by a model
The source article does not present In the Weights as a scientific measure of importance. It describes a site that purports to measure model recall, and it includes skepticism. AI critic Anthony Moser scoffed that this is “literally the same as asking 13 chatbots to tell you about yourself.”
That critique points to the central tension. In the Weights is compelling because it makes AI memory visible, but the output is still generated by models that can vary, misidentify people, or produce hallucinations. A high score can feel validating. A strange answer can feel revealing for a very different reason.
The site’s design also appears to help its spread. The source article describes it as having a cute, Nintendo-inspired retro design. That presentation makes a technical idea easier to play with: instead of reading about model behavior in the abstract, users can type in names and watch the rankings change.
What comes next for In the Weights
Dimson said he plans to investigate several questions further. He wants to look at why different models in the same series return different results, which models are biased toward different types of people, and which people “should have a Wikipedia article but don’t.”
Those questions move the project beyond simple curiosity. If models recall some people more strongly than others, the differences may say something about how information is represented inside AI systems. If one model family gives a different picture than another, that could matter for anyone whose reputation is increasingly filtered through AI answers.
For now, In the Weights works because it compresses a big shift into one readable number. Search visibility is no longer the only public mirror. Chatbot recall is becoming another place where people look for proof that they are known.