Wispr Flow: Dictation That Works EverywhereStop typing. Start speaking. 4x faster.
Promoted
Facts...No Bullsh*t
Stop BS in real-time with AI that fact-checks as you listen
78 followers
Stop BS in real-time with AI that fact-checks as you listen
78 followers
Most fact-checkers make you wait — copy text, paste links, hope for results. By then, the damage is done. We built fact-checking that happens while you listen. Turn it on during a podcast, meeting, or call. Our AI detects verifiable claims in real-time and instantly checks them against reliable sources. What's different: Truly real-time — verified as words are spoken Works with any audio source Clear TRUE/FALSE verdicts with sources Know what's true the moment it's said.
Facts...No Bullsh*t
Facts...No Bullsh*t
@mike_fitzbaxter The pipeline has two independent stages:
Claim detection: analyzes the transcribed audio and extracts only verifiable factual claims. It's explicitly instructed to ignore opinions, questions, and small talk. It doesn't decide what's true or false — it just identifies what can be checked.
Fact-checking: each claim is verified via live web search, not from a static knowledge base. It returns a verdict (TRUE / FALSE / UNCERTAIN), a brief explanation, and a direct link to the source so you can verify it yourself.
On neutrality and bias:
There's no user customization of verdicts. Everyone gets the same result for the same claim — there's no "check only things I disagree with" option, so no echo chamber by design.
The tool doesn't hold positions on topics. It delegates verification entirely to external web sources.
The UNCERTAIN verdict exists specifically for cases where evidence is conflicting or insufficient. It doesn't force a binary true/false when reality is more nuanced.
Honest limitations:
Bias can enter at two levels: what the llm considers a "verifiable claim" and which sources the llm surfaces from the web. No fact-checking system, human or automated, is fully immune to this.
The quality of results depends on the quality and availability of sources on that specific topic.
Facts...No Bullsh*t
@hex_miller_bakewell Hi, I have fixed the language detection. Now it s in the same language of the audio. Thanks for the feedback