Former OpenAI science lead launches Warp Science to automate wet labs with robots
I keep thinking about those sensor gloves. Chemistry has always been a craft as much as a science, the way you know a precipitate is ready by how it swirls, the feel of a pipette that's about to drip. If Warp Science can capture that embodied knowledge and hand it to a robot, we're not just automating tasks. We're bottling the muscle memory of an entire field.
What interests me is what happens to the people doing the training. There's a strange inversion here: the human becomes the teacher, the demonstrator, the oracle, while the robot becomes the tireless student. I'd want to know how they handle the messy edge cases, the contamination events, the "this batch just feels wrong" moments that no protocol ever captures. That's where the real value lives, and also where the training data gets thin.
I'm cautiously optimistic. We've seen enough failed lab automation to be skeptical, but the sensor-rich approach feels genuinely different. The question is whether the robots will ever develop the intuition to notice when something is off without being told. That might be the last frontier, and it might be a very long frontier indeed.