Use Case Title:

ByesAlert/Gaurd - Train not just your body, but your mind too

Description:

This Chrome feature detects words in articles with biased connotations, like "appalling display of" or "barbaric actions," which express the author's opinion and evoke negative emotions. An AI highlights these sentences, offers labels for vocabulary types (e.g., hyperbole, imagery), measures connotation (positive, neutral, negative), and provides a history of analysis from various sources for a broader perspective.

Tutorial for Use and Best Practices:

  • Encourage users to research claims independently.
  • Promote critical thinking and fact-checking.
  • Utilize Limiter for educational purposes to teach students about emotional manipulation in online content.

Impacts on Learning:

Educational Impact:

  • Limiter aids in teaching students about online bias and emotional manipulation.
  • Encourage users to research claims independently.
  • Promotes critical thinking, fact-checking and media literacy.
  • Educational Use Example: Employ Limiter to teach students about online emotional manipulation, even in news articles from major publications (e.g., link that uses biased language "brutally")

Limitations and Ethical Considerations:

Potential future problems to look out and prepare for

  • Limiter's effectiveness may vary based on the AI's ability to detect connotations accurately (Databse updated through community interaction and polls)
  • It cannot completely eliminate bias but instead focuses on raising awareness.
  • Users might rely solely on the tool instead of developing critical thinking skills.
  • While Limiter promotes awareness, it should not restrict free speech or diverse opinions.
  • AI Limitations: Beware of potential false positives and false negatives when highlighting connotations. This could result from the AI's lack of updated information. Google's BARD may be more suitable for current event analysis (vs. ChatGPT)

Possible solution to the above listed problem

  • De-limiter could utilise a voting system similar to that found in SponsorBLOCK to help give news articles ratings and see community votings on words flagged by the AI - Tool could include comments and remarks from other schools to help students understand how perspectives vary from different countries on the same topic and how taking apersons, upbringing and history into account is important for a diverse discussion.

Inspiration

I have always been someone to criticize others. Its something I have always believed as being integral to having healthy debates. Especiylly nowadays the importance of critical thinking is more important than ever as "journalists" try to alter the connotations of news articles to increase their ad revenue and bolster their pockets with more money. However, with impartial news articles free from bias one can fully construct a rational plane of thought, with all necessary viewpoints, ye in this day and age that is but wishful thinking.

Implementation

Mix of integration in google classroom/Moodle/Microsoft Teams with an chrome add on for use on news articles itself

What it does

This Chrome feature will identify words in articles that have specific meanings and could be used to sway the reader. One instance would be if an article's headline said anything like "appalling display of..." or "barbaric actions..." These statements all express the author's opinions while deviating from a true account of the events. They also have the effect of stirring up hate and violent imagery in the reader's head. It's as if the authors don't even want the reader to exercise independent thought.

These words take away the freedom of thought from the reader implanting them with emotions that the writer intended to gain even more attention. But, an AI language model could be able to identify these kinds of sentences and draw the reader's attention to them by highlighting the subject of the sentence with tiny bars (additional features: labelling englisch vocab type used, ie. hyperbole, imagery etc; positive, neutral and negative connotation meter; history of analysis from other schools for the same article - broadens perspective)

Furthermore, emphasis should be put on "alleged" claims and "accusations" through colour coding and small text bubbles, to let the reader know that these are not hard bound facts but instead arguments without a complete foundation. In fact, the add-on could make a list of all the claims made on that website, and below a list of allegations and claims, with links to other websites with similar info (for cross-checking purposes). These could be displayed in a small side panel (Potential use for Google Bard)

These are only some of the possibilities ByesAlert can bring to fruition allowing students to learn how they can get influenced online in a subtle manner, by playing to their emotions. This way not only will our classroom technology advanced, but so will be the students operating said technologies.

Educational Impact:

-Limiter aids in teaching students about online bias and emotional manipulation. -Encourage users to research claims independently. -Promotes critical thinking, fact-checking and media literacy. -Educational Use Example: Employ Limiter to teach students about online emotional manipulation, even in news articles from major publications (e.g., link that uses biased language "brutally")

Potential future problems to look out and prepare for

-Limiter's effectiveness may vary based on the AI's ability to detect connotations accurately (Databse updated through community interaction and polls) -It cannot completely eliminate bias but instead focuses on raising awareness. -Users might rely solely on the tool instead of developing critical thinking skills. -While Limiter promotes awareness, it should not restrict free speech or diverse opinions. -AI Limitations: Beware of potential false positives and false negatives when highlighting connotations. This could result from the AI's lack of updated information. Google's BARD may be more suitable for current event analysis (vs. ChatGPT)

Possible solution to the above listed problem

  • ByesAlert could utilise a voting system similar to that found in SponsorBLOCK to help give news articles ratings and see community votings on words flagged by the AI
  • Tool could include comments and remarks from other schools to help students understand how perspectives vary from different countries on the same topic and how taking apersons, upbringing and history into account is important for a diverse discussion

Built With

  • chat-gpt
  • chromium
  • google-bard
  • google-classroom
  • microsoft-teams
  • moodle
Share this project:

Updates