The Minutes of Dylan Patel on cost reduction https://minutesof.com/dylan-patel/on/cost-reduction/ Everything Dylan Patel has said on cost reduction: 2 verbatim quotes between March 2025 and July 2026, each with a timestamp and a link to the recording… en Mon, 31 Aug 2026 13:10:11 +0000 Patel argues layering KV cache offload and multi-token prediction on open-source engines drastically cuts infe https://minutesof.com/q/6f64d719-d15d-49df-847e-a57b86cfaaa1/ https://minutesof.com/q/6f64d719-d15d-49df-847e-a57b86cfaaa1/ “So if you take the, you know, just open source inference engines off the shelf, that gets you a certain level of cost.” — RAISE Summit Thu, 16 Jul 2026 16:44:57 +0000 cost reduction Patel states same-quality AI model output became 1200x cheaper than GPT-3 through successive model releases. https://minutesof.com/q/baf6aa96-f58f-4dc9-aa6e-f3c668d9d92e/ https://minutesof.com/q/baf6aa96-f58f-4dc9-aa6e-f3c668d9d92e/ “Anthropic released new models. Google released new models. Meta released new models. Right? And then now it is 1200x times it's 1200x cheaper for the same output.” — MedBricks Webcast Thu, 27 Mar 2025 13:55:08 +0000 cost reduction