This new tool could give artists an edge over AI

However artists are the canary within the coal mine. Their combat belongs to anybody who has ever posted something they care about on-line. Our private information, social media posts, music lyrics, information articles, fiction, even our faces—something that’s freely out there on-line might find yourself in an AI mannequin perpetually with out our understanding about it. 

Instruments like Nightshade could possibly be a primary step in tipping the ability steadiness again to us. 

Deeper Studying

How Meta and AI firms recruited hanging actors to coach AI

Earlier this 12 months, an organization referred to as Realeyes ran an “emotion examine.” It recruited actors after which captured audio and video information of their voices, faces, and actions, which it fed into an AI database. That database is getting used to assist practice digital avatars for Meta. The mission coincided with Hollywood’s historic strikes. With the business at a standstill, the larger-than-usual variety of out-of-work actors might have been a boon for Meta and Realeyes: right here was a brand new pool of “trainers”—and information factors—completely suited to educating their AI to look extra human. 

Who owns your face: Many actors throughout the business fear that AI—very like the fashions described within the emotion examine—could possibly be used to interchange them, whether or not or not their actual faces are copied. Read more from Eileen Guo here.

Bits and Bytes

How China plans to evaluate generative AI security
The Chinese language authorities has a brand new draft doc that proposes detailed guidelines for how one can decide whether or not a generative AI mannequin is problematic. Our China tech author Zeyi Yang unpacks it for us. (MIT Technology Review) 

AI chatbots can guess your private info from what you kind
New analysis has discovered that enormous language fashions are glorious at guessing individuals’s non-public info from chats. This could possibly be used to supercharge profiling for commercials, for instance. (Wired) 

OpenAI claims its new device can detect photographs by DALL-E with 99% accuracy
OpenAI executives say the corporate is growing the device after main AI firms made a voluntary pledge to the White Home to develop watermarks and different detection mechanisms for AI-generated content material. Google introduced its watermarking tool in August. (Bloomberg)

AI fashions fail miserably in transparency
When Stanford College examined how clear massive language fashions are, it discovered that the top-scoring mannequin, Meta’s LLaMA 2, solely scored 54 out of 100. Rising opacity is a worrying development in AI. AI fashions are going to have large societal affect, and we want extra visibility into them to have the ability to maintain them accountable. (Stanford) 

A university scholar constructed an AI system to learn 2,000-year-old Roman scrolls
How enjoyable! A 21-year-old pc science main developed an AI program to decipher historic Roman scrolls that have been broken by a volcanic eruption within the 12 months 79. This system was capable of detect a few dozen letters, which consultants translated into the phrase “porphyras”—historic Greek for purple.  (The Washington Post)