The Shift from SEO to Weights Optimization

Traditional search engine optimization is losing relevance as information consumption pivots from web indexing to model parameter recall. The emergence of “In the Weights”—a tool that quantifies how deeply an individual or brand is embedded within an LLM’s internal knowledge base—marks the beginning of a new “Answer Engine Optimization” (AEO) era.

What Happened

Launched in June 2026 by former OpenAI engineers Thomas Dimson and Joey Flynn, In the Weights queries major LLMs (Grok, Gemini, GPT, Claude, Llama) to generate recall-based scores for individuals. By analyzing the consistency and detail of model outputs without external web search, the tool provides a metric of “AI notoriety.” It essentially treats an LLM’s internal weights as the ultimate authority on market relevance, bypassing traditional backlink-based metrics.

Why It Matters

First-order: Operators can no longer rely solely on PageRank to measure digital footprint. If your target demographic relies on AI chatbots for discovery, your presence within model training data is now more critical than your SERP ranking.

Second-order: This triggers a shift in PR and marketing spend. Agencies will pivot from traditional link-building to “AI-seeding” strategies designed to ensure brands are represented correctly within the high-density training data of foundation models.

Third-order: The long-term decay of Google’s traditional search dominance creates a power vacuum in brand discovery. We expect a rise in specialized “LLM-visibility” audits, similar to the SEO consultancy boom of the 2010s, as companies scramble to improve their “recall score” across fragmented models.

What To Watch

  • AEO Tooling Proliferation: Look for third-party platforms that offer “LLM sentiment” tracking and citation monitoring across disparate models.
  • Model-Specific Optimization: As models diverge, expect companies to develop unique strategies to ensure they are represented in specialized models (e.g., coding-focused vs. consumer-focused models).
  • Data Poisoning Defenses: As brand equity becomes tied to model weights, companies will prioritize securing their proprietary data to prevent AI “hallucinations” that damage their score.