Yoon et al. 2026 redraws rather than breaks the computation-first boundary of autonomous science: the discovery is biological, but it was made by surveying protein-cluster databases—work conducted entirely on the computer, the very condition under which Lu et al. 2026 built their end-to-end system—so the autonomy frontier tracks the digital experimental loop, not the discipline.
E23 shows the biological discovery (novel RT family) was made by computationally surveying protein-cluster databases, and E3 states Lu et al. built their end-to-end system focusing on ML science because experiments occur entirely on the computer, directly implying the autonomy frontier tracks the digital experimental loop rather than the discipline.
Written by Kimi K3 via Ollama Cloud · checked by GLM-5.3 via Ollama Cloud · 31 Oct, 04:33
Source chain
AEvery quote below was checked, without a model, to appear verbatim in its source.
- 01
“We deployed Claude Code instances to survey reverse transcriptase (RT) loci across 1.9 billion protein clusters”
Abstract · no page number · evidence E23
- 02
“We focus on machine learning science, as experiments typically occur entirely on the computer.”
Full-text passage · no page number · evidence E3
Other systems focused on analysing pre-collected datasets to find new insights10,17,18. However, with the recent advent of powerful and general foundation models, the role of AI has expanded to include assisting with a wider array of research activities. For example, LLMs now help with generating new hypotheses19–23, writing literature reviews24,25 and coding experiments26–29. Despite these advances in automating individual components, a system that autonomously navigates the entire research life cycle—from conception to publication—has remained out of reach until now. This paper introduces The AI Scientist, a pipeline that achieves the vision of full end-to-end automation of the scientific process. The AI Scientist uses existing foundation models to perform ideation, literature search, experiment planning and implementation, result analysis, manuscript writing, and peer review to produce complete, new papers. We focus on machine learning science, as experiments typically occur entirely on the computer. A central challenge in developing such a system is automatically evaluating the quality of its scientific output at scale.
Passage read from www.ebi.ac.uk, which may be a preprint rather than the published version.