Michael Dempsey

This is one of the best pieces I've read on AI in some time. "How can anyone stop being fascinated for long enough to be angry?"

Perhaps the biggest constraint surrounding further intelligent autonomy in robotics is training data. Enter: Using diffusion models to generate long-tail and hard to gather robotics training data. Great thread by @xf1280

Another great and tangible thread on how robotics can move forward from this time of optimism productively. Continue to believe this is a very slept on category for foundation models & LLMs as everyone looks at AI + SaaS.

One of the more thought-provoking ML papers I've read in awhile. Researchers built a model to determine Remaining Life (years till you die) based on a photo. Their model predicted within 8 years (SOTA age detection models are within 2-3 years).

Great thread again from @hausman_k on utilizing LLMs within robotics