Explore a 25-minute IEEE conference talk on optimal streaming approximations for Boolean Max 2-CSPs and Max k-SAT. Delve into constraint satisfaction problems in the streaming model, examining how trivial random sampling proves optimal for Max-CUT. Investigate the Max-DICUT streaming model, including a 2/5-approximation and a novel random sampling with bias approach. Analyze the relationship between total bias and cut value, and study streaming lower bounds through communication complexity. Examine the Distributional Boolean Hidden Partition (DBHP) problem and its reduction to Max-DICUT. Gain insights into future research directions in this field of computational complexity and approximation algorithms.