nivekney

I think so, the benchmark is on a coding dataset (SPEED-Bench).

This is outrageous. Someone go create a Polymarket.

User data integrity definitely should be a concern. It's also known that regulations is being outpaced, so the cost of being/using frontier products is a double-edged sword for sure.

"We"? Who "We"?

Map-reduce as a pattern might be on its way back. Hear me out. High localization wins even when coverage is not super great -- just map shards of the corpus and reduce the learnings. Rinse and repeat, do as many rounds of map and reduce to traverse the corpus until converge. This can also work well when the cluster is combined with different agents, they are tasks equally by prompts anyway.

Just when I think I cannot see more emdashes, there are more emdashes.

Wait a second, Snapchat impacted AGAIN? It was impacted during the last GCP outage.

I see Rust, I like

It's based on ParlayLib, which is for shared-memory multicore machines. Highly suspect that they moved the algorithms on to distributed systems.