# The F1 Principle: Sources & Citations | SAVRN Source: https://savrn.com/sources/the-f1-principle Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- [All research sources](https://savrn.com/sources) Sources & Citations # The F1 Principle 20 primary sources 5 categories We publish the receipts [The essay The F1 Principle](https://savrn.com/blog/the-f1-principle) A ## The numbers behind the analogy 480 horsepower at 7 · 4 megajoules per lap · Formula 1 Ford lists its current 5.0-liter Mustang V8 at 480 horsepower at 7,150 rpm. 480 horsepower at 7,150 rpm [View source](https://performanceparts.ford.com/part/M-6007-M50D) The other recovered energy under braking and put up to 120 kW back into the drivetrain, 4 megajoules per lap. 4 megajoules per lap [View source](https://www.formula1.com/en/latest/article/from-smaller-cars-to-a-bigger-budget-cap-12-rule-changes-you-need-to-know-in.56uUTFhB0z5j3iZfhC0rGP) Formula 1’s own explainer says the new unit “retains the 1.6-litre turbocharged V6 engine (with a few tweaks) but deletes the MGU-H,” while the electric motor moves to “350kW to the rear wheels, up from 120kW,” and the electrical share of… Formula 1 [View source](https://www.formula1.com/en/latest/article/2026-regulations-explained-all-you-need-to-know-about-f1s-new-power-units.14jfv7a36905uDJDdNyfQd) The combustion side actually gets smaller: the FIA’s technical explainer has the internal combustion engine dropping “from between 550-560kW to 400kW,” with recovery rising to around 8.5 megajoules per lap. FIA via Formula 1 [View source](https://www.formula1.com/en/latest/article/explained-2026-power-unit-regulations-fia.68izKQ2tn1voQPWvgLVMXN) Formula 1 says the 2026 units “will still provide over 1,000 horsepower” (Formula 1). Read that again. Formula 1 [View source](https://www.formula1.com/en/latest/article/more-efficient-less-fuel-and-carbon-net-zero-7-things-you-need-to-know-about.ZhtzvU3cPCv8QO7jtFxQR) Mercedes AMG High Performance Powertrains reports its F1 unit “reaching more than 50% thermal efficiency”. Mercedes [View source](https://www.mercedesamgf1.com/team/location/brixworth) When Mercedes first broke that barrier on the dyno in 2017, the 2014 unit had been at 44 percent. Autosport [View source](https://www.autosport.com/f1/news/mercedes-f1-engine-hits-remarkable-efficiency-target-on-dyno-4995489/4995489/) A typical road-car engine turns roughly 35 percent of its fuel into motion. Motor Authority [View source](https://www.motorauthority.com/news/1112999_mercedes-amg-f1-engine-achieves-50-percent-thermal-efficiency) B ## Displacement is not the answer Princeton HAL · arXiv 2605.23950 · Stanford HAI On the GAIA agent benchmark, the same model, Claude Sonnet 4.5, scores 30.91% in one scaffold and 74.55% in another. Princeton HAL [View source](https://hal.cs.princeton.edu/gaia) A 2026 paper that held the model fixed and changed only the scaffold found the same pattern across benchmarks and concluded that “harness-induced variance can substantially exceed model-induced variance”. arXiv 2605.23950 [View source](https://arxiv.org/html/2605.23950v1) Meanwhile the engines themselves are converging: Stanford’s AI Index measured the gap between the best closed model and the best open-weight model at 8.04% in January 2024 and 1.70% by February 2025. Stanford HAI [View source](https://hai.stanford.edu/ai-index/2025-ai-index-report/technical-performance) C ## The harness is where the race is won Mercedes · AWS · Formula 1 Mercedes says its car carries “over 250 sensors” and produces about “30 megabytes per lap,” more than a terabyte per car per weekend, over 17 separate data buses. Mercedes [View source](https://www.mercedesamgf1.com/news/feature-data-and-electronics-in-f1-explained) AWS, which runs F1’s data platform, puts it at 300 sensors per car “generating more than 1.1 million data points per second” (AWS). The strategy is not improvised either. AWS [View source](https://aws.amazon.com/sports/f1/) F1’s own insiders’ guide has a strategist explaining that “Strategy is 98 per cent preparation,” that the simulations run on Saturday night, and that “Everything is based on statistics”. Formula 1 [View source](https://www.formula1.com/en/latest/article/watch-the-insiders-guide-to-f1-strategists.6VT3445ugVd0bh85Pw5slO) We have Monte Carlo, we have gen AI applications” (AWS Executive Insights). And the pit crew rehearses until the stop is muscle memory. AWS Executive Insights [View source](https://aws.amazon.com/executive-insights/podcast/ai-and-data-strategy-in-f1-with-ruth-buscombe/) The record is 1.80 seconds, set by McLaren at the Qatar Grand Prix on October 8, 2023. Guinness World Records [View source](https://www.guinnessworldrecords.com/world-records/108666-fastest-formula-one-pit-stop) D ## What the V8 operators are actually buying LangChain · McKinsey LangChain’s State of Agent Engineering found 94% of teams have observability tooling and only 77.2% actually evaluate their agents’ output. LangChain [View source](https://www.langchain.com/state-of-agent-engineering) McKinsey reports that more than two-thirds of high-performing companies name data, not the model, as the primary obstacle to scaling AI. McKinsey [View source](https://www.mckinsey.com/capabilities/mckinsey-technology/our-insights/ai-data-readiness-the-key-to-scaling-impact) E ## Detailed plans win. Big plans lose. Anthropic · Simon Willison Instead, they were building with simple, composable patterns”. Anthropic [View source](https://www.anthropic.com/engineering/building-effective-agents) Simon Willison’s definition of a coding agent is even shorter: “LLM + system prompt + tools in a loop”. Simon Willison [View source](https://simonwillison.net/guides/agentic-engineering-patterns/how-coding-agents-work/) These are the primary sources behind our research. Have a better one, or spot an error? [Tell us](https://savrn.com/contact?topic=sources) — we’ll correct it.