DEV Community

Ben Dixon profile picture

Ben Dixon

I test how reliable consumer AI tools are on real questions — the same prompts through ChatGPT, Claude, Perplexity and Gemini, every answer graded against the source. I publish where they get it wrong

Location UK Joined Joined on  Personal website https://dixon.ai github website
I scored 12 AI models on whether they admit they don't know. My own token limit nearly broke the test.
bendixon profile

I scored 12 AI models on whether they admit they don't know. My own token limit nearly broke the test.

#ai #llm #machinelearning #testing
Comments
3 min read

Want to connect with Ben Dixon?

Create an account to connect with Ben Dixon. You can also sign in below to proceed if you already have an account.

Create Account
Already have an account? Sign in
loading...