I wanted to start a broader mistakes and lessons thread around General AI Robotics.
A useful angle is security, but it also connects with useful tools. Different tools and workflows can produce very different results even when the goal is similar.
It would be interesting to compare practical...