Nathan Labenz and Prakash Narayanan lead this AI:AM highlights episode with a live, hosts-only exploration of Anthropic’s “global workspace” paper, including the J-space and J-lens claims about readable concepts inside language models and the limits of wha...