AI Search Engine for Digital Legacy Launched by Ex-OpenAI Staff

In 2026, former OpenAI employees launched In the Weights, a free website designed to answer a profound question: do you 'live forever in the super intelligence'? This new AI-centric vanity search engine measures how well AI models recall your personal information, tapping into a deep human desire for digital permanence, as Bitcoin World observed. Our digital identity once resided solely in external web presence. Now, this tool reveals it also exists as an internal recall within AI models, creating a new, unseen dimension of self. Consequently, individuals will increasingly seek to understand and influence this AI-centric digital footprint, fundamentally shifting how we manage our online identity.

What is an AI-Centric Vanity Search?

In the Weights acts as an AI-centric vanity search. It measures how Large Language Models (LLMs) recall individuals without relying on external web searches, according to Bitcoin World. The tool assesses if an AI model can identify, describe, and contextualize a person internally, Startup Fortune reported. A new, distinct form of digital identity, one rooted in AI's internal memory rather than public web presence, is created. This unindexed layer remains largely invisible and unmanageable, a challenge In the Weights itself highlights. The implication is profound: our digital selves now exist in a realm we cannot easily access or control.

Benchmarking AI Models for Personal Data

In the Weights tests leading AI models: GPT-5.5, Claude Opus 4.8, Gemini, Grok, and Llama. It queries them about a person and assigns a 'strength score,' Startup Fortune noted. This benchmarking establishes a new standard for gauging how deeply an individual's information is embedded within cutting-edge AI. It implies that AI's 'understanding' is a synthesized, persistent internal representation, one that could evolve independently of our public personas.

How AI Models Are Queried for Personal Information

In the Weights directly queries models like Grok, Gemini, GPT versions, Claude, and Llama. It asks, 'Who is [name]?' and requests up to 10 descriptive results with confidence levels, according to Bitcoin World. This direct approach highlights the critical need to understand AI's internal data representation. While The Tech Buzz describes it as showing a 'digital footprint in AI training data,' implying a direct view, Startup Fortune clarifies the tool measures AI recall 'without performing an external search.' This tension is crucial: users are not seeing raw training data, but rather a synthesized, potentially distorted, and uneditable internal representation. The implication is a loss of agency over one's own narrative within these powerful systems.

The desire to 'live forever in the super intelligence' is a powerful draw for tools like In the Weights, as Bitcoin World notes. An ancient human need for enduring legacy is now projected onto the digital realm. Traditional digital legacy focused on public archives and external web presence. In the Weights introduces a new layer: private AI recall. The concept of digital legacy fundamentally shifts from public records to internal AI knowledge. For AI companies, this new identity layer presents an urgent challenge: how to provide transparency and control when models internalize and recall personal information without external search. Opaque digital identities demand new user management tools, or risk alienating those whose digital selves are now beyond their grasp.