Instagram uses AI to combat bullying

Picture and video-sharing app Instagram announced it will take on pervasive Internet bullying on its platform. New head of Instagram Adam Mosseri wrote in a blog post that the team aims to spread kindness, and will do so by utilizing AI and machine learning that can identify bullying and harassment. The technology optically scans photos that include bullying tactics like comparing attacks on a person's appearance or character, ranking and rating images, and threats to someone's wellbeing and health. The photos will then be sent to a human team of community moderators for review.

In May of this year, Instagram was already starting its fight against online harassment by introducing a tool that blocks offensive comments. Users could turn on a "Hide Offensive Comments" option that could hide explicit sexual content or harassment for both posts and live video. 

AI and machine learning can be used for good in a multitude of ways. In the healthcare realm, the technology is making waves with predicting types of lung cancer. A  Stanford study describes teaching computers to teach themselves the assessment of lung cancer tissue. This approach is quantitative, objective, and is able to pinpoint tiny differences between normal tissue and those with cancer.

The same can be said for early skin cancer detection—AI can help see minute changes in the concentration and distribution of eumelanin and hemoglobin, two indicators of melanoma. Alzheimer's is seeing a breakthrough as well, with Verge Genomics automating its drug discovery engine to rapidly develop treatments for diseases like Alzheimer's, Parkinson's, and ALS. The key is in the company's algorithms on human genomic data.

Beyond healthcare, AI can find good in the unconventional. Cisco Meraki's Product Manager for Camera Systems George Bentinck relayed a story about finding a bicycle thief in their offices' bike room through Meraki's AI-powered smart camera, the MV12. The AI in the camera and its cloud technology allowed Bentinck to quickly find the camera footage from the bike room via the Meraki dashboard. From there, he could search the footage based on motion, key-in on the exact bicycle that went missing, and the camera automatically queried what happened to that bike based on how it physically moved.

Cisco is doing a ton more with AI and machine learning, including embedding it into products like Encrypted Traffic Analytics and Webex Assistant. Learn more about that here.