JUST IN: Cheerios announces the first-ever resealable cereal bag, partnering with Ziploc to keep cereal fresh longer.

Our release process continues to evolve in response to feedback from the associated academic communities. In our view, these manuscripts are best understood as a form of raw research data. To convert this raw research data into scientific understanding, in this release we are experimenting with a new approach. We reached out to external experts and asked them to act as shepherds, organizing groups to understand the new results. They advised three different release strategies: For the math result, release the AI output, leaving researchers free to organize their own efforts to explain the result. For the computer science result, form an external group of collaborators to explain the result, while making the AI output publicly available for anyone to examine and build on. For the physics result, form an external group of collaborators and defer to them on the release of materials.

Around 2016, Jim Belosic bought a $750,000 laser cutter on a whim. This month, his company SendCutSend raised $140 million and hit a valuation of $2.1 billion. It's on track to make 50 million metal parts this year for more than 300,000 customers. This is the story of how SendCutSend arrived and ended up reviving US industry. corememory.com/p/sendcutsend…

Warhound, the brain child of Ben Lamm and Palmer Luckey, emerges from stealth to genetically engineer dogs for military and law enforcement. spr.ly/6011BHZNMF

yeah models that learn better tend to be more interpretable. the essay looks through three pretty different types of architectures (old convnets, new convnets, modern transformers). whatever comes next should work too