arising from Pithan & C.-M. Hubner Nature Geoscience https://doi.org/10.1038/s41561-026-02078-y (2026) The degree of Arctic amplification (AA) varies strongly with the analysis period, being ~1.9 over the period 1970–2005 but ~3.8 for 1985–2020. In our original paper1 we showed that this period-dependent AA reflects natural variability superimposed on a steady threefold forced AA. In particular, the interdecadal Pacific oscillation modulates global temperature, while an Arctic mode modulates Arctic temperature. Pithan and Hubner2 argue that the Arctic mode identified in our paper is an artefact arising from applying linear detrending to a period with little Arctic warming before 1985. Here, we rebut their argument and present additional evidence for the existence of the Arctic mode. We emphasize that the period before 1985 should be included because global temperature had already warmed notably and the muted Arctic warming indicates the effect of natural variability. We show that the Arctic mode is consistently identified from empirical orthogonal function (EOF) analysis when the forced response is not assumed to be linear, when alternative longer and shorter periods are considered, and when analysis is restricted to the Arctic region with area weighting. The Arctic mode differs notably from the long-term warming trend in both spatial pattern...

Technology companies say that their services can sniff out AI-generated writing, but their tools vary in technique and quality.

The laser-accessible nuclear transition in the thorium-229 isotope has been identified as a promising candidate for the realization of an optical nuclear clock. Such a nuclear clock might rival or out...

This week, Nature is publishing a paper1 with, perhaps, an unusually underwhelming finding at its heart: that a particular technique failed to improve how artificial neural networks learn. It was not these findings that our editors found noteworthy. Rather, it was how the work was done, which is in fact the main focus of the paper. AI is saving time and money in research — but at what cost? The result was created using The AI Scientist engineered by researchers at Tokyo-based company Sakana AI, details of which were first made available2 as a preprint in 2024. This artificial-intelligence system represents an effort to automate the scientific process entirely, from performing literature review and conceiving an idea to executing experiments and writing them up. As detailed in the Nature paper, The AI Scientist was able to follow this process and generate a research paper about its (negative) result; the work passed the first round of peer review for submissions to a workshop at a major machine-learning conference. AI research assistants have since proliferated, with technology firms Google, OpenAI and Anthropic all trialling ways to automate research. Although their outputs have been limited and rarely innovative so far, the effects of...