The video addresses a common pain point with AI chatbots (inaccuracy) and offers a practical, actionable solution. The "Truth Prompt" concept is intriguing, and the creator leverages a clear demonstration and a direct call to action (commenting for the prompt) to drive engagement and potential virality.
Summary
This video explains how to use a custom prompt in Claude AI to ensure it provides truthful and verified information. The "Truth Prompt," developed by MIT researchers, prevents Claude from fabricating information or guessing, and instead forces it to verify its logic and facts before responding.
ProLocked — included with ProLocked
Transcript, structure and on-screen text
5 beats, a 313-word transcript and 29 lines of on-screen text — the parts you need to write your own version.
The hook, the summary, why it worked and what it sells stay open on every video, signed in or not.
Original caption
Comment "Truth" and I'll send you the prompt 👇 Claude lies to you 17% of the time. Here's one prompt that forces it to always tell the truth. It's called the Truth Prompt, developed by researchers at MIT to stop Claude from guessing, filling in gaps, and confidently telling you things that aren't even true. Here's how to install it: Go to Claude and click your name in the bottom left Go to Settings → Instructions for Claude Paste it in and hit Save Now Claude will: • Verify and check its logic, facts, completeness, and bias before every answer • Reflect on its own answer • Retry the output if the truth confidence meter is below 0.8 One paste. Saves forever. Now Claude stops lying and starts telling you the truth. Comment "Truth" and I'll send you the prompt plus every other prompt I use every day. #claude #ai #claudeprompts #aitools #claudetruth
More from @maverickgpt
Their other posts in the index, biggest breakout first.