Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript.
Daily briefing: How ancient humans bred and traded dogs
The diversity of present-day dogs seems to have emerged thousands of years ago. Plus, an artificial intelligence model taught itself basic physics and a tiny ‘robot’ that swims through blood vessels.
The diversity of modern-day dog breeds might have emerged thousands of years ago. An analysis of ancient skulls reveals a distinctive dog skull shape — a shortened snout and widened face, compared to wolves — first seen in nearly 11,000-year-old fossils. Not long after, researchers say, there was a boom in the variety of dog skull shapes and sizes, capturing more than half of the diversity in today’s breeds. An analysis of the genomes of ancient dogs also suggests that ancient humans valued distinctive features, took their canine companions with them when they moved around and traded dogs with useful characteristics.
A remote-controlled ‘robot’ the size of a grain of sand can swim through blood vessels to deliver drugs before dissolving into the body. The system uses tiny gelatine beads filled with a drug and nanoparticles of iron oxide, which allows its movement to be guided with a magnetic field. The ‘microrobots’, which have been tested in pigs and sheep, could allow doctors to deliver small doses of drugs to specific sites with less risk of toxic off-target effects.
A system called AI-Newton, after being fed simulated experimental data, can autonomously ‘discover’ key physics principles, such as Newton’s second law. The model mimics the human scientific process by incrementally building a knowledge base of concepts and laws, says physicist Yan-Qing Ma, who helped to develop the system. It is far from making truly autonomous discoveries, but Ma thinks it could help develop future AI systems that use real-world data to discover new general laws of physics.
Researchers have revealed the details of the technology behind AlphaProof, a tool designed to prove mathematical statements that did well at problems from last year’s International Mathematical Olympiad (IMO), a competition that sets tough maths problems for gifted high-school students. The achievement kicked off “a space race of sorts”, with other tools reaching gold-medal scores on this year’s IMO problems, writes proof-engineering researcher Talia Ringer. She shares insights gained by trying out AlphaProof herself.
AlphaProof is just one example of how artificial intelligence tools are reshaping the way mathematical research is conducted, write mathematicians Bartosz Naskręcki and Ken Ono. They think it’s unlikely that AI systems will solve deep problems that require the development of new theories. But they do expect that AI integration will transform mathematical practice. It could boost the need to formalize theorems and proofs into machine-readable formats, and make human creativity and insight even more valuable.
Andrew Robinson’s pick of the top five science books to read this week includes an itemized study of ‘technofossils’ — the remnants of modern civilization — and an illustrated guide to the subtleties of typography.
Researchers have detected what they say is the strongest evidence yet of a coronal mass ejection (CME) — a huge expulsion of magnetized plasma — coming from a star other than our Sun. This changes everything for the study of CMEs, astronomer and study co-author David Konijn tells the Nature Podcast. While previous assumptions had to rely on the star we know best, “now we actually have this stellar data point”, he says. Next, “we're going to search through this entire data set very deliberately, very carefully to maybe find some more dimmer bursts”, says Konijn.
Vicarious trauma — psychological distress in individuals who are exposed to other people’s trauma through work — is an occupational hazard for some researchers, says gender-based violence researcher Millan AbiNader. She is one of many advocating for better workplace policies to ensure that such academics receive appropriate emotional support. (Nature | 5 min read)
doi: https://doi.org/10.1038/d41586-025-03771-5
In today’s penguin-search puzzle, Leif Penguinson is enjoying the tamarind forest in the Plaine des Tamarins, Réunion Island. Can you find the penguin?
The answer will be in Monday’s e-mail, all thanks to Briefing photo editor and penguin wrangler Tom Houghton.
This newsletter is always evolving — tell us what you think! Please send your feedback to briefing@nature.com.
• Nature Briefing: Careers — insights, advice and award-winning journalism to help you optimize your working life
• Nature Briefing: Microbiology — the most abundant living entities on our planet — microorganisms — and the role they play in health, the environment and food systems
