This week - an update on Neuralink's progress; San Francisco police propose using robots capable of "deadly force"; how AI will transform the animation industry; and more!
More Than A Human
This week, Neuralink hosted an event to announce what the company was working in. Elon Musk said that Neuralink has begun submitting paperwork for a human clinical trial to the FDA, and hopes to implant a Neuralink device in a patient in six months. The team also demonstrated how its surgical robot inserts electrode threads, using a dummy filled with a gelatinous “brain proxy” material. The full live-stream recording (all 2 hours and 48 minutes) is available on YouTube.
OpenAI released ChatGPT - a chatbot that can "answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests", according to OpenAI. MIT Technology Review checked these claims and it looks like they do generally hold, but there is still a lot of work to be done for these models to be useful and trustworthy.
This is an interesting application of deepfakes. BBC used face-swapping AI to protect the identity of protesters in Hong Kong while maintaining their facial movements and emotional expressions.
This article looks at the future of the animation industry and how AI-generated art is going to impact it. It argues that within the next 5 years, AI will dominate animation and we are already seeing a glimpse of what is to come. But this should not be seen as a death of artistic expression in animation. "By removing that bottleneck, writers and visionaries will be freed to craft and imagine even better stories. The next ten to twenty years will probably feature an era of human artistic abundance".
Sofia Crespo shares her experiments in using AI to generate never-seen-before creatures. What started as an art project with 2D images of jellyfish-like creatures turned into 3D animated models of new insects with their own, also AI-generated, scientific names.
The San Francisco police department has proposed that it be allowed to use robots with “deadly force” while responding to incidents. The police department already has a collection of robots, many of which can be easily weaponised. The SFPD proposal would allow these robots to kill people “when risk of loss of life to members of the public or officers is imminent and outweighs any other force option available to SFPD”.
Researchers from the University of Zurich invited three drone racing champions to race against their AI pilots in two categories. In the first category, the AI had access to 36 cameras positioned above the flight path. In the second, the AI worked only with onboard cameras. In both categories, the human pilots won (for now).
Wired invited robotics professor Henny Admoni to answer the internet's burning questions about robots, from why we like building humanoid robots to questions about self-driving cars to could some fictional robots become reality.