> "So, though now I have started working again in cognition (but in isolation), I can’t avoid seeing the problem coming."

Quoted from just above the heading: "What can be done?"

It appears that he continues to work on the simulation of human cognition.