Mike Fridman

Love seeing someone go straight at the boring but painful bottlenecks. Chunking rarely gets attention until it breaks something downstream. 40x faster + flat memory is a pretty compelling combo — especially for teams pushing larger corpora. Out of curiosity, what pushed you to rebuild this in Rust instead of optimizing within Python first? Always interesting to see where people decide “this...

Spent some time exploring Winion. What surprised me wasn’t just the geopolitics angle — it’s how simple it is to spin up a private market. Created a test event in under a minute and watched probabilities move once a few people joined. That dynamic shift in confidence is actually fascinating. Feels like a hybrid between serious forecasting and social game theory. The AI-assisted probability...