How about using AI against cyberbullying on social media?
Burak SERPEN
17 replies
After using AI support in our workplace, now it's time to use it for our digital security! 🤖
An AI-based social media analyzing tool Socialays is launching its free social media cyberbullying prevention app Socialays: Social Media Guard very soon on the Product Hunt! 🚀
Check out Socialays's upcoming page and subscribe to support us on the launch day! 🙌
https://www.producthunt.com/upcoming/socialays-social-media-guard/
Replies
Satish Kumar Veluri@satish_kumar_veluri
I was wondering how do we get to know that AI is not involved in cyberbullying?
Share
Delphi — Digital Clone Studio
What are your thoughts on instagrams built in detections? Stuff like the restricted words feature etc. I have an insta account with 102k followers and I heavily rely on blocking certain keywords in comments.
Socialays: Social Media Guard
@philipsnyder Instagram's comment filtering works well (If You can filter out all the words that are considered hate speech and bullying)...
Socialays's NLP-powered social media sentiment analyzing features always expand its ''keywords'' to detect and categorize your social media comments with AI; such as positive, negative, question, ad content, spam, hate speech, and profanity! 🤬❌🤖
Check out Socialays's free social media cyberbullying prevention and comment analyzing app on Product Hunt 🙌
https://www.producthunt.com/upco...
I think this needs AI + human intelligence based solutions
Socialays: Social Media Guard
@chetan_natesh Socialays = Machine level speed, human level accuracy 🚀
AI prevents cyberbullying from further spreading. – Algorithms determine precisely who the bully, the victim, and the bystanders are in each.
Socialays: Social Media Guard
@todd_osborn Like Socialays does for the social media comments..
I think it will be a large step forward and result in a lot less false positives. For now, most social media uses keyword detection, and a lot of comments that are fine get removed because they contained a specific keyword.
With the sentiment analysis power of AI, I think we'll see a lot more people being more open and free to talk about controversial topics from an analytical frame of view rather than using a slew of censored language to get past the keyword detectors.
I think it will also help with detecting cyberbullying in non-English languages, since I've noticed keyword detection is relatively weak for non-english words.
But on the flip side, I think it might also be used for malicious purposes censoring free speech in authoritarian countries.
It's all very exciting.
Is your product powered by GPT-3?
It's the best on the market right now, but there are open source alternatives such as Bloom, which has more languages (46) it can detect. I think detecting hate speech for non-english languages is a huge market right now considering how much of our architecture is built on English.
Currently, we're working on an API for Bloom that runs on the cloud much cheaper than GPT-3. If you're curious and have more questions, you can join our discord: https://discord.com/invite/DsJXx...
Socialays: Social Media Guard
@richard_gao2 Thank you so much for your comment and sharing your valuable insights, Richard! I'll do more research about GPT-3 and Bloom...
Flamme AI - The Couples App
Wonderful idea! I think our generation really needs more tools like this to help identify cyberbullies.