Beyond the Search Bar: Are You Digitally Immortalized in the Weights of AI?

For decades, the ultimate barometer of one’s cultural relevance or public standing was the "Google search." To be searchable was to exist; to be at the top of the results page was to be recognized. However, as the landscape of information retrieval shifts away from traditional blue-link search engines toward generative artificial intelligence, that metric is losing its luster. In an era where chatbots serve as the primary interface for knowledge, a new project—In the Weights—is emerging to challenge the old guard of vanity metrics, asking a provocative question: If you aren’t in the model, do you really exist?

The Paradigm Shift: From Search Engines to Neural Networks

The internet is currently undergoing a structural transformation. As Google and other search providers grapple with the integration of Large Language Models (LLMs) into their core products, the way users consume information is changing. We no longer just "search" for a subject; we ask a conversational agent to synthesize it for us.

This evolution has rendered the traditional "Google vanity search" somewhat obsolete. When users want to know about a public figure, a journalist, or an author, they are increasingly likely to query ChatGPT, Claude, or Gemini rather than scrolling through a list of indexed web pages. This shift has created a new, invisible hierarchy: the internal knowledge base of AI models.

In the Weights, a new platform created by former OpenAI engineers Thomas Dimson and Joey Flynn, aims to quantify this hierarchy. The site measures the extent to which an individual has been "encoded" into the neural parameters—the numerical weights—that dictate how an AI model generates its output.

Chronology of a Digital Identity Project

The genesis of In the Weights can be traced back to the post-OpenAI ambitions of Dimson and Flynn. After their design startup, Global Illumination, was acquired by OpenAI in 2023, the duo eventually departed, looking for a way to spark their creative instincts in a new direction.

The conceptual spark was ignited by a blend of technical curiosity and existential philosophy. Dimson notes that the project was heavily influenced by Terry Bisson’s classic short story, "They’re Made Out of Meat," which explores the nature of consciousness and the physical composition of intelligence. This, coupled with a viral blog post by Max Leiter regarding AI weights, led the team to wonder about the "human" data buried deep within the massive matrices of modern AI.

By mid-2026, the project moved from a theoretical concept to a functional web tool. The developers began querying a wide array of models—including Grok, Gemini, various iterations of GPT, Claude, and Llama—with a standardized prompt: "Who is [Name]? Give up to 10 results, each with a short description and confidence." The resulting data is then clustered, analyzed, and assigned a "strength score," effectively creating a leaderboard of digital importance.

Supporting Data: The Anatomy of a Score

The mechanics of the site are designed to be both intuitive and revealing. When a name is submitted, the system pings multiple models to determine their "knowledge" of that individual without resorting to real-time web browsing. This ensures that the results reflect what the model has actually "learned" during its training phase, rather than what it can simply scrape from the live web.

The Scoring Metric

The "strength score" is the heart of the project. A high score suggests that an individual’s identity, biography, and impact are deeply ingrained in the model’s training data.

• The Top Tier: Current leaders include pop-culture icons like Macaulay Culkin and historical figures like Luciano Pavarotti, both of whom command strength scores approaching 1,000.
• The "Humble" Reality: For many professionals, the results are sobering. For instance, tech journalist Anthony Ha received a score of 641, placing him in the top 6% of all tracked entities. While statistically significant, it serves as a reminder that even those who are "known" in their respective industries are often just minor data points in the vast, multi-billion parameter landscape of a model like GPT-5.4.

The Hallucination Factor

One of the most critical aspects of In the Weights is its ability to expose the limitations and "hallucinations" of AI. When models are asked about individuals they aren’t entirely sure of, they often struggle to provide a cohesive narrative. For example, GPT-5.4 Mini’s attempt to identify Anthony Ha as an "ambiguous name form" that could refer to multiple people demonstrates the propensity for AI to "fill in the gaps" when its confidence interval is low.

Official Responses and The Philosophy of "Being"

In conversations regarding the project, Thomas Dimson has been candid about the project’s intent. "Google vanity searches are the wrong objective in 2026 as more traffic moves to LLMs," he explained. The goal, according to Dimson, is to confront the reality that for millions of people, their life’s work, history, and existence are now "encoded somehow in a bunch of floating-point numbers inside the AI brain."

The reception has been, by his own admission, "insane." While the developers initially envisioned the site as a "mild curiosity," it has clearly struck a nerve. The public’s desire to see if they "live forever" in the super-intelligence of the future is a powerful motivator. The site’s design, which features a charming, Nintendo-inspired aesthetic, masks the somewhat unsettling reality of what is actually being measured: the cold, calculated extraction of human identity into digital parameters.

Not everyone, however, is convinced of the site’s profound significance. AI critic Anthony Moser has dismissed the project as "literally the same as asking 13 chatbots to tell you about yourself." From this perspective, In the Weights is a mirror reflecting the inherent biases and training data limitations of LLMs, rather than a definitive measure of human worth or legacy.

Implications for the Future of Personal Branding

The rise of In the Weights signals a massive shift in how we think about digital presence. If the goal of the 2010s was Search Engine Optimization (SEO)—making sure Google’s crawlers could find your website—the goal of the late 2020s may well be "LLM Optimization."

The Bias of the Model

Dimson and his team plan to expand the project to explore why specific models in the same series yield different results. This raises critical questions about algorithmic bias. Are certain demographics or professional groups systematically excluded from the "weights"? Does a model’s training data favor Western-centric narratives, and if so, what happens to the global digital footprint of those who fall outside that data set?

The "Wikipedia" Paradox

One of the most intriguing avenues for future research is identifying which individuals should have a Wikipedia article but don’t, yet still maintain a high score in an LLM. This suggests that AI models are building a secondary, perhaps more robust, encyclopedia of human existence that exists independently of the traditional, human-edited internet.

Is This Digital Immortality?

The ultimate question remains: Does being "in the weights" constitute a form of immortality? While it is certainly a form of digital preservation, it is fragile. Unlike a book or a monument, the data inside a neural network is subject to the volatility of the model’s architecture. When a model is updated, pruned, or replaced, the "weights" change. The version of "you" that exists in GPT-4 might vanish or shift when GPT-6 is released.

In the Weights serves as a playful yet sobering reminder that our digital lives are no longer stored in static files, but in the fluid, evolving synapses of machines. Whether or not this makes us immortal, it certainly makes us part of the architecture of the future. As we move forward, the question of who we are will increasingly be defined not by what we say, but by what the machine has decided to remember about us.