Rules that require all AI-generated content to be watermarked are impossible to enforce, and it’s also highly possible that watermarks could end up doing the opposite of what they’re supposed to do, Leufer says. For one thing, in open-source systems, watermarking and provenance techniques can be removed by bad actors. This is because everyone has access to the model’s source code, so specific users can simply remove any techniques they don’t want.
If only the biggest companies or most popular proprietary platforms offer watermarks on their AI-generated content, then the absence of a watermark could come to signify that content is not AI generated, says Leufer.
“Enforcing watermarking on all the content that you can enforce it on would actually lend credibility to the most harmful stuff that’s coming from the systems that we can’t intervene in,” he says.
I asked Leufer if there are any promising approaches he sees out there that give him hope. He paused to think and finally suggested looking at the bigger picture. Deepfakes are just another symptom of the problems we have had with information and disinformation on social media, he said: “This could be the thing that tips the scales to really do something about regulating these platforms and drives a push to really allow for public understanding and transparency.”
Deeper Learning
Watch this robot as it learns to stitch up wounds
An AI-trained surgical robot that can make a few stitches on its own is a small step toward systems that can aid surgeons with such repetitive tasks. A video taken by researchers at the University of California, Berkeley, shows the two-armed robot completing six stitches in a row on a simple wound in imitation skin, passing the needle through the tissue and from one robotic arm to the other while maintaining tension on the thread.
A helping hand: Though many doctors today get help from robots for procedures ranging from hernia repairs to coronary bypasses, those are used to assist surgeons, not replace them. This new research marks progress toward robots that can operate more autonomously on very intricate, complicated tasks like suturing. The lessons learned in its development could also be useful in other fields of robotics. Read more from James O’Donnell here.
Bits and Bytes
Wikimedia’s CTO: In the age of AI, human contributors still matter
Selena Deckelmann argues that in this era of machine-generated content, Wikipedia becomes even more valuable. (MIT Technology Review)