The more you model the memory of a system, the less you have to model the structure. It’s a sort of anti-markovian approach. Markov = maximal-state, 1-step memory. Narrative is minimal-state, maximal memory. Kinda like the big data principle (simpler algorithms, more data)
Conversation
This Tweet was deleted by the Tweet author. Learn more
Replying to
Currently juggling between CPU load and buffer latency while producing audio so I feel ya
1

