AutoScientia — Notes from the lab

AutoScientia — Notes from the lab https://autoscientia.com/blog/ Writing about automating the scientific method: the loop, the measurement, the engineering and the horizon. en © 2026 AutoScientia Tue, 8 Sep 2026 09:00:00 +0200 From machine learning to the grand challenges https://autoscientia.com/blog/from-ml-to-grand-challenges/ https://autoscientia.com/blog/from-ml-to-grand-challenges/ Tue, 8 Sep 2026 09:00:00 +0200 Roadmap Why we start where the measurement is honest, what has to be true before the loop is pointed at medicine, energy or water, and the sequencing we are willing to be judged against. Five ways an autonomous loop fools itself https://autoscientia.com/blog/how-the-loop-fools-itself/ https://autoscientia.com/blog/how-the-loop-fools-itself/ Tue, 30 Jun 2026 09:00:00 +0200 Rigor Metric drift, survivorship bias in the search history, favourable seeds, self-graded objectives and the slow move of the target — and how each one looks from the inside. Acting as our own first customer https://autoscientia.com/blog/our-own-first-customer/ https://autoscientia.com/blog/our-own-first-customer/ Tue, 21 Apr 2026 09:00:00 +0200 Strategy Why the loop is pointed at our own stack before anyone else's, which parts of machine learning are worth automating first, and the failure modes of self-optimisation. Running a thousand experiments without losing one https://autoscientia.com/blog/orchestrating-parallel-experiments/ https://autoscientia.com/blog/orchestrating-parallel-experiments/ Tue, 17 Feb 2026 09:00:00 +0100 Engineering Scheduling, preemption, checkpointing and budget ceilings — the unglamorous systems work that decides whether parallelism is real. The loop only works if you can measure it https://autoscientia.com/blog/measurement-first/ https://autoscientia.com/blog/measurement-first/ Thu, 11 Dec 2025 09:00:00 +0100 Measurement A frozen baseline, held-out objectives, repeated seeds and a failure taxonomy. The harness is the product, and most of our engineering goes into it. What we mean by “autonomous experimentation” https://autoscientia.com/blog/autonomous-experimentation-defined/ https://autoscientia.com/blog/autonomous-experimentation-defined/ Tue, 9 Sep 2025 09:00:00 +0200 Fundamentals A precise definition: four stages, what the machine owns, what stays with people, and the three tests a cycle has to pass before it deserves the word autonomous. Why science is still hand-cranked https://autoscientia.com/blog/why-science-is-hand-cranked/ https://autoscientia.com/blog/why-science-is-hand-cranked/ Tue, 24 Jun 2025 09:00:00 +0200 Thesis The bottleneck is not ideas or instruments. It is a sequential loop that no one can run fast enough by hand — the observation the company was founded on.