Facebook PixelReal-time misinformation detection app that lights up when nonsense is detected in conversations
Create newCreate new

Real-time misinformation detection app that lights up when nonsense is detected in conversations

Image credit: sungkd / oneplus forums

Darko Savic
Darko Savic Oct 24, 2021
Please leave the feedback on this idea

Is it original or innovative?


Is it feasible?


Is it targeting an unsolved problem?


Is it concisely described?

Bounty for the best solution

Provide a bounty for the best solution

Bounties attract serious brainpower to the challenge.

Currency *
Who gets the Bounty *
An app that listens to conversations and fact-checks the statements in real-time. Upon detection of nonsense, a notification goes off.
  • Protect people from the spread of misinformation.
How it works
Let me start by saying that the technology might not be quite there yet, but we should be pretty close. This was in 2011.
  1. Two people are having a conversation and one of them decides that it would be a good idea to fact-check everything they say. They turn on the app on one of their phones.
  2. The app listens to speech and converts it to text in real-time.
  3. Sentences that don't sound right to GPT3 are ignored. We can assume that the software did not properly understand/convert the speech to text. As the technology evolves the number of ignored sentences should be reduced.
  4. Sentences that seem ok when convereted to text are fact-checked online.
  5. If common misconceptions/misinformation is detected, a discreet notification beep/light goes off.
  6. The speaker can elaborate on what they said to provide more context and see if this passes without triggering the notification.
  7. Either way, the conversation continues.
  8. At the end of the conversation the entire transcription is sent to both parties. The misconceptions are marked and resources to proper information linked on the side.
  9. If the software made a mistake in identifying misconceptions, any of the conversation participants can set it straight (provide citations and quote reputable sources). This gets reviewed by the app moderators/developers that continuously work on improvements.
Different modes
Depending on the purpose of the conversation, the software can be configured to detect different things. For example people can opt to be defended from logical fallacies, manipulative sales, etc.
Creative contributions

Detecting personal "white" lies

Povilas S
Povilas S Oct 29, 2021
This is not to approach total honesty between people, although it could perhaps be used for that as well if some people desired this and would mutually consent to it. What I mean instead is using this for particular people who are known for mixing fantasies into their personal stories to the level that others can't tell what's true and what's not about their life/past experiences. Some readers who dealt with those type of people will know what I'm talking about. I have two friends who are like that.
The problem is that they are good people otherwise, but they like to fantasize, it's just part of their character and this quality becomes tiring after communicating longer with them, they don't admit their lies if you try to challenge some "facts", but instead will try to justify it further or present some more lukewarm alternative version of it. We could call this the Munchausen effect. There's a psychological condition called the Munchausen syndrome, although it's explicitly about exaggerating/imagining health problems, therefore a term Munchausen effect could perhaps be used for the cases that I've described.
I understand it would be complicated for a proposed app to debunk personal lies (in comparison to universally known facts) as well, but I think it could nevertheless be useful at least in some particular cases. Say someone claims that they worked in this extravagant company a few years ago, the app instantly checks the register of companies that were operating at that time in a specific location and finds out that such a company never existed.
I wonder how such people would react if the conversation with them happened in the presence of a proposed app. If the app was generally approved as effective they would probably avoid lying in the first place. The app could even be used silently in the presence of a particular person and compose a profile of him/her indicating a general trustworthiness level of their claims. The app couldn't confirm/debunk everything, but from the things that it could, it could quickly guess how prone to lying that person is.
Please leave the feedback on this idea

Add your creative contribution

0 / 200

Added via the text editor

Sign up or


Guest sign up

* Indicates a required field

By using this platform you agree to our terms of service and privacy policy.

General comments

Michaela D
Michaela D2 years ago
I would love to use that during politicians' speeches.
Please leave the feedback on this idea
Spook Louw
Spook Louw2 years ago
Theoretically, the software should also be able to check legal documents for you? Even at a basic level, it would be an excellent way to provide cheap or free legal care to those who can't afford it.
Please leave the feedback on this idea