In the Weights: The AI Vanity Search Tool You Didn’t Know You Needed
In the Weights is a new tool that shows how well AI models know your name or brand. Here's what it does, why it matters, and how to use it.
LUMIEN4 min read
A new tool called In the Weights lets you check how well your name or brand is represented inside AI language models, essentially a vanity search built for the AI era. Rather than Googling yourself to see where you rank in search results, you can now query how much weight AI systems give to your presence. The concept, covered by TechCrunch in June 2026, points to a growing concern for businesses: if an AI model doesn't know who you are, you may not exist for a growing slice of users.
What happened
A tool called In the Weights has launched with a straightforward premise: type in your name or brand, and find out how prominent you are inside AI language models. TechCrunch reported on it in June 2026, framing it as the AI-era equivalent of the old habit of Googling yourself.
The core idea is that AI models are trained on large datasets of text from the web. The more a name, company, or concept appears in that training data, the more confidently a model can discuss it. In the Weights tries to surface a score that reflects that representation.
Why it matters
For years, businesses have tracked their Google rankings as a proxy for visibility. That logic made sense when most discovery happened through search. But more users are now getting answers directly from AI assistants, skipping the results page entirely.
If an AI model has thin or inaccurate knowledge of your brand, a user asking about your product category might never hear your name mentioned. There is no page-two equivalent here. You either show up in the model’s answer or you don’t.
This creates a practical gap for small and mid-sized businesses. Large brands with years of press coverage, Wikipedia entries, and broad web mentions are well represented in training data almost by default. Smaller operators who rely on local SEO or paid ads have had little way to know where they stand inside these models.
In the Weights attempts to make that gap visible. A score gives you something concrete to react to, even if the field of improving that score is still very young.
Our take
We’re cautiously interested in this, not because the score itself is definitive, but because it puts a number on a real problem our clients are starting to ask about.
The honest truth is that “AI search optimization” is not yet a mature discipline. Nobody fully controls what a model retains from its training data, and there is no Google Search Console equivalent that tells you exactly why a model underweights your brand. So treat any score from a tool like this as a directional signal, not a report card.
That said, the underlying question is worth taking seriously right now:
- Is your business mentioned in credible, crawlable sources that AI training pipelines are likely to include?
- Do third-party sites describe what you do accurately, or is most of your web presence locked in social media posts and PDFs?
- Does your site’s content clearly explain your products, location, and category in plain language?
These are all things that help both traditional SEO and AI model representation. The tactics overlap more than the hype suggests.
We’d also flag one reason for skepticism: tools that produce a single score for a complex, opaque system can create false confidence or false panic. A low score might reflect that your industry is niche, not that you’ve done something wrong. A high score doesn’t mean AI tools are recommending you positively.
What to do about it
If you want to improve how AI models represent your brand, start with the basics before chasing any new tool’s score:
- Get covered by text-based, publicly indexed sources. Trade publications, local news, industry directories, and Wikipedia (where appropriate) are the kinds of sources that end up in training data.
- Keep your own site’s language clear and factual. Describe what you do, who you serve, and where you operate in plain sentences. Avoid jargon that only insiders recognize.
- Audit third-party descriptions of your brand. Review sites, partner pages, and press mentions often contain outdated or inaccurate descriptions. Correct what you can.
- Run your own spot checks. Ask ChatGPT, Claude, or Gemini about your brand today. Note what they get right and wrong. That’s your baseline.
Check your In the Weights score for a starting point, but put more effort into the content and coverage that actually shapes what models learn.
