Zoom has updated its ToS to allow training AI on user content without the ability to opt-out. The line that pisses me off is that Zoom has a “perpetual, worldwide, non-exclusive, royalty-free, sublicensable, and transferable license” to redistribute, publish, access, use, store, transmit, review, disclose, preserve, extract, modify, reproduce, share, use, display, copy, distribute, translate, transcribe, create derivative works, and process Customer Content. (ᔥ Stackdiary; ↬ Abeba Birhane)

Researchers at NVidia are showing off Perfusion, a text-to-image model they say is 100KB in size and takes four minutes to train. What could possibly go wrong? (ᔥ Hackaday)

Mongabay on breadfruit’s “biodiversatility”: According to recent research, the increased temperatures of climate change will widen breadfruit’s range, especially in Sub-Saharan Africa […] making this a reliable crop for places struggling with poverty and food security.

John D. Cook’s three advantages of non-AI models: It’s hard to imagine doing anything like Midjourney’s image generation without neural networks. The same is true of ChatGPT’s text generation. But a lot of business tasks do not require AI, and in fact would be better off not using AI.