Researchers at the University of Sydney demonstrated that an AI network made of nanowires is more efficient than computer based AI because it can learn and adapt in real time.
Researchers at the University of Chicago developed a tool called Nightshade that can be used to poison images used to train AI models, in order to interfere with the art generated by those models.
Nvidia Research have developed an AI agent using ChatGPT that can teach robots complex skills autonomously. The skills include catching and throwing balls, opening cabinets, and using scissors.
Biotech firm Klick Labs developed a 10-second smartphone voice test that uses AI to detect type 2 diabetes, with an accuracy of 89% for women and 86% for men.
Google's Project Green Light uses AI and the vast amount of Google Maps traffic data to identify efficiencies in traffic schedules in cities that can deliver 30% less stopping and 10% less traffic emissions.
The Toyota Research Institute released details of their work towards developing a Large Behaviour Model (LBM) for robotics. The AI model learned 60 dexterous behaviours autonomously following a manual demonstration of each behaviour using physical simulation.
