Tweet 1 Every LLM call burns GPU cycles on tokens that never needed to run. Padding. Boilerplate. Irrelevant context. I built SuperCompress β€” a tiny CPU policy that cuts 65% of tokens before inference. Open source. MIT. Free tier. supercompress.vercel.app Tweet 2 The problem is worse than most people realize. At ~50M agent turns/day: β†’ 100B tokens wasted daily β†’ 24K GPU hours β†’ 1,526 tons COβ‚‚ β†’ 6.5M L cooling water We're burning through resources...