1. Mozilla conducted an investigation that accused YouTube’s recommendation algorithms of powering harmful content. It found that France and Germany were particularly affected, as well as other non-English speaking countries. The report’s conclusion is that YT’s algorithms are designed to harm and mislead people. 71% of videos reported as harmful content were automatically recommended by the platform. The main types of harmful content identified were videos containing misinformation, violence, hate speech or scams. More info here.
2. The European Parliament adopted on July 6, 2021 a final version of a new regulation providing a legal framework for technology companies to voluntarily monitor human communications to detect and report material depicting the sexual abuse of minors. This temporary measure is intended to allow providers of electronic communications services to scan and report private Internet messages containing child sexual abuse material. The legislation also allows companies to use approved technology to detect child grooming techniques. However, the regulation has been criticized as being too intrusive and allowing mass monitoring of private communications. The scanning of private conversations will be done through automated content recognition tools using artificial intelligence, but under human supervision. Read more here.
3. Researchers at MIT have created a programmable fiber that contains memory, temperature sensors, and a trained neural network program to infer physical activity. The material is thin and flexible and can be sewn into fabrics and washed. Read more here.
4. The research team from MIT and Massachusetts General Hospital has developed advanced machine learning software that can help anesthesiologists choose the optimal drug dose. The AI can adjust the type of anesthesia used by using an algorithm that evaluates brain activity. Read more here.
5. Researchers at Columbia Engineering have created an artificial intelligence that learns to predict human behavior from video footage. The AI has analyzed thousands of hours of videos, games, and TV series, allowing the system to learn to predict hundreds of actions, from a handshake to a fist bump. The technique can teach computers to evaluate certain situations and make decisions based on that. Read more here.
6. The team of researchers at Nanyang Technological University in Singapore (NTU Singapore) has developed millimeter-sized robots that can be controlled by magnetic fields to perform very efficient tasks. In the future, they could be used in biomedicine and manufacturing. Read more here.
7. IBM’s fully autonomous, artificial intelligence-powered Mayflower ship has set sail on an unmanned voyage from Plymouth, UK to Plymouth, USA. Mayflower 2.0 is a state-of-the-art vessel equipped with the latest technological advances. AI computing on board enables the ship to conduct scientific research while sailing in harsh ocean conditions – often without communications. The AI had to be specially trained to ignore seagulls, as they can appear to be large objects and lead to the Mayflower taking unnecessary actions to maneuver around them. Read more here.
8. The White House has launched a website with official updates on the U.S. government’s National Artificial Intelligence Initiative. Click here to visit the site.
Author: Agata Konieczna
