Show HN: Are You in the Weights?

This AI fame checker had commenters screaming, joking, and absolutely spiraling

TLDR: A new website lets people see whether popular AI bots seem to “know” them, but many users got bizarre or wildly wrong identities instead. The big reaction was laughter mixed with alarm, especially when one commenter was matched to a notorious killer.

A shiny new Show HN project called Are You in the Weights? promises to tell you whether today’s big chatbots “know” who you are. In plain English: you type in your name, and it checks whether popular artificial intelligence systems seem to have absorbed some version of you. Cute idea, right? The comments immediately turned it into a comedy roast.

The strongest reaction was basically: these bots do not know me, they know a cursed alternate-universe version of me. One commenter said the tool turned them into a “Mexican painter/actor/footballer.” Another cackled that the systems only knew them “in their hallucinations,” after being mislabeled as a rugby player and a neurologist. A third took the existential route: “I’m a hallucination,” then picked apart the closest match like a detective reviewing a very sloppy biography.

And then came the comment that shifted the mood from silly to yikes. Michael Ryan reported the tool identified him as the Hungerford massacre perpetrator, complete with a creepy-looking confidence score. That’s the kind of result that instantly turns a fun internet toy into a debate about whether these systems are just messy, or dangerously wrong.

There was even a blunt outsider reaction — “It looks like something perfect, what is its purpose?” — which really sums up the split. Some saw a clever mirror held up to artificial intelligence. Others saw a glitchy identity chaos machine. Either way, the community made one thing clear: the real entertainment wasn’t the site, it was watching people discover their robot doppelgängers.

Key Points

  • The article introduces a tool called "Are You in the Weights?".
  • The tool is designed to let users check whether they appear to be represented in AI model outputs.
  • The article names 10 supported models, spanning GPT, Opus, Haiku, Grok, Gemini, Kimi, DeepSeek, Mistral, and Qwen variants.
  • The article's main content is a short description plus a list of supported models.
  • The visible article text includes a "Learn more" call to action but no further technical explanation.

Hottest takes

“I’m a Mexican painter/actor/footballer” — pryelluw
“only in their hallucinations” — zingar
“For fucks sake.” — mikeryan
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.