Headline: In the Weights — The AI-Powered Vanity Search Revolution
In an era where Google search no longer reigns supreme as the definitive source of personal information, a new tool has emerged to measure a different kind of digital presence: how well artificial intelligence models remember you. In the Weights, launched by former OpenAI designers Thomas Dimson and Joey Flynn, offers users a unique vanity score derived from querying multiple large language models (LLMs).
Key Fact 1: The Concept of Being "In the Weights"
The term "weights" refers to the numerical parameters that shape an AI model's training and output. In the Weights purports to measure how well a model can recall someone without using external tools like web search. The website states: "Being in the weights means your existence was deemed important in the process of creating superhuman artificial intelligence." This concept taps into the growing fascination with digital immortality and how AI encodes human knowledge.
Key Fact 2: How the Scoring Works
To generate a score, In the Weights queries a range of models—including Grok, Gemini, multiple versions of GPT, Claude, Llama, and lesser-known models—with a prompt like: "Who is <name>? Give up to 10 results, each with a short description and confidence." The system then clusters similar descriptions and assigns a strength score. For instance, a tech blogger received a score of 641, placing them in the top 6% of names. The leaderboard shows Macaulay Culkin and Luciano Pavarotti tied near the top with scores around 988.
Key Fact 3: The Founders' Background and Motivation
Thomas Dimson and Joey Flynn previously worked at OpenAI, joining through the acquisition of their design startup Global Illumination. Dimson told TechCrunch that they built In the Weights to "get the creative juices flowing again" after leaving the AI giant. He was inspired by the observation that Google vanity searches are outdated in 2026 as more traffic shifts to LLMs, and by the idea that "so many lives are encoded somehow in a bunch of floating point numbers inside the AI brain." The site's direction was further sealed by a tongue-in-cheek blog post riffing on AI weights and Terry Bisson's short story "They're Made out of Meat."
Key Fact 4: Reception and Criticism
The tool has garnered significant attention, with Dimson noting that reception was "insane" and that it struck a nerve among users wanting to see if they 'live forever in the super intelligence.' However, critics like AI commentator Anthony Moser have dismissed it as "literally the same as asking 13 chatbots to tell you about yourself." The site also highlights potential hallucinations—for example, GPT-5.4 Mini claimed that Anthony Ha is an "ambiguous name form that could refer to multiple people with the initials A.H.A."—adding an element of humor and unpredictability.
Key Fact 5: The Role of Hallucinations and Model Bias
In the Weights not only provides a score but also shows which models returned which answers, highlighting discrepancies. This transparency allows users to see biases across models—some may favor certain demographic groups or well-known figures. Dimson plans to investigate why different models in the same series return different results, which models are biased toward different types of people, and which individuals "should have a Wikipedia article but don't." This data could shed light on the inherent biases in training data and model architecture.
Expanded Context: The Shift from Traditional Search to AI
The decline of Google as the go-to source for personal information is well-documented. As chatbots like ChatGPT, Gemini, and Claude become primary interfaces for information retrieval, the question of how these models represent individuals becomes increasingly important. In the Weights fills a niche by providing a quantifiable measure of representation, albeit one that is highly dependent on the model's training data. This shift mirrors the broader transition from search engines to generative AI, raising questions about accuracy, fairness, and the nature of digital identity.
Background on Vanity Search and Its Evolution
Vanity searching—searching for one's own name—has been a staple of internet culture since the early days of Google. In the past, a high Google ranking meant visibility and influence. Today, being mentioned in a chatbot's response can be equally or more significant, especially as AI-generated answers replace traditional search results. In the Weights capitalizes on this trend by offering a new metric: the strength score. However, the tool's reliance on model memory, which is inherently flawed and biased, raises questions about its validity as a measure of importance.
Technical Details and Model Selection
The platform queries a diverse set of models to ensure broad coverage. By including both well-known models (GPT-4, GPT-5, Claude 3.5) and more obscure ones, the tool aims to capture a range of perspectives. The clustering algorithm groups similar descriptions to avoid redundancy and assigns a strength score based on frequency and consistency of recall. This approach, while innovative, is not without limitations. Models may hallucinate details or fail to recall individuals with low representation in training data, leading to skewed results.
User Experience and Design
In the Weights features a retro, Nintendo-inspired design that adds a touch of nostalgia to the experience. Users can search for any name and instantly view their score, a leaderboard, and detailed breakdowns per model. The visual simplicity contrasts with the complex underlying computation, making the tool accessible to a broad audience. The design also encourages sharing and competition, driving viral growth. Users can compare their scores with friends or celebrities, fueling the desire for higher rankings.
Future Directions and Implications
Dimson has outlined plans to explore deeper questions about model biases and the representation of lesser-known individuals. The tool could potentially be used to audit AI models for underrepresentation of certain groups or to identify gaps in training data. As AI becomes more pervasive, services like In the Weights may become standard for monitoring digital identity. However, the founders have not announced plans for monetization or long-term sustainability, leaving the project as more of a creative experiment than a commercial venture.
Comparisons to Other Tools
While there are other tools that scrape the web for personal information, In the Weights is unique in its focus on AI model recall. Services like Google Alerts or Mention monitor web mentions, but they don't assess internal model knowledge. Meanwhile, platforms like Wikipedia provide canonical sources, but they are curated by humans. In the Weights bridges the gap by directly probing the brains of AI systems, offering a glimpse into what these models 'know' about individuals. This makes it a fascinating, albeit imperfect, gauge of digital presence in the age of generative AI.
Ethical Considerations
The ability to query multiple models for personal information raises privacy concerns. While In the Weights only returns information that is already publicly available in training data, it aggregates it in a new way. Users may be surprised to see what models 'know' about them, especially if the information is outdated or hallucinated. The tool also amplifies existing biases in training data, potentially reinforcing stereotypes or overrepresenting certain demographics. The founders have not addressed these ethical issues in depth, but they remain an important part of the conversation.
Impact on AI Research
By providing a crowdsourced dataset of model recalls, In the Weights could become a valuable resource for AI researchers studying grounding, hallucinations, and bias. The tool effectively turns vanity searching into a form of model evaluation. Researchers could use the data to compare how different models treat the same entity, identifying weaknesses or inconsistencies. This could lead to improvements in model training and data curation, making future AI systems more accurate and equitable.
Personal Narratives and Anecdotes
Many users have shared their scores on social media, creating a viral phenomenon. Some express delight at high rankings, while others lament low scores. The emotional investment in these numbers highlights our desire for recognition, even from machines. For example, one user with a modest score of 450 was pleased to be remembered at all, while another with a score of 950 joked about being an AI celebrity. These reactions underscore the psychological impact of digital visibility in an increasingly AI-mediated world.
Source: TechCrunch News