![\"AI](\"https:\/\/media.nature.com\/lw767\/magazine-assets\/d41586-026-01199-z\/d41586-026-01199-z_52283248.png?as=webp)
![\"\"](\"https:\/\/media.nature.com\/w400\/magazine-assets\/d41586-026-01199-z\/d41586-026-01199-z_51572570.png\"\/)
<\/p>\nHow AI agents will change research: a scientist\u2019s guide<\/p>\n
<\/a><\/article>\n<\/p>\n
The boom in AI-related science papers is \u201cnot surprising\u201d, says computer scientist Arvind Narayanan at Princeton University in New Jersey, who was not involved with creating the index. But it\u2019s not clear, he says, whether the rise of AI use is productive for science. \u201cWhether or not this explosive growth is meaningful is hotly debated,\u201d Narayanan says. \u201cMy view is that it is happening too fast, without giving scientific norms time to adjust, and so the quality of research has taken a nosedive.\u201d<\/p>\n
Gil says there isn\u2019t much evidence yet that AI is improving scientists\u2019 productivity. \u201cThe studies are limited,\u201d she says. But, she adds, scientists \u201ccan\u2019t live without it. If you took AI away from them, there would be a riot. So it must be helping in some way.\u201d<\/p>\n
Top models<\/h2>\n
As uptake has grown, so has the number of AI platforms available to researchers. This past year saw the emergence of many science foundation models, including the first one for astronomy, AION-1, which was trained on more than 200 million celestial objects. This training helps it to classify galaxies or estimate their properties. \u201cWhen I talked to scientists in 2024 and said \u2018There\u2019s foundation models for science\u2019, scientists would not know what that means. They didn\u2019t know they existed. I think we have seen that really advance very quickly,\u201d says Gil.<\/p>\n<\/div>\n