New AI algorithms are 95% better at showing how the universe changes over time
New AI algorithms named GAME are 95% more effective at modeling how the universe evolves over time, offering astrophysicists a clearer view of cosmic changes. These algorithms use a ensemble-based approach inspired by natural selection to improve accuracy in interpreting observational data. The advancement could challenge or refine the current standard cosmological model by better capturing subtle dynamics like rates of change.
- ▪GAME stands for Genetic Algorithms with Marginalised Ensembles and uses multiple algorithms instead of a single one.
- ▪The technique improves the accuracy of measuring cosmic derivatives, which describe how quickly aspects of the universe are changing.
- ▪Traditional genetic algorithms struggle with nonobservable quantities and rates of change, but GAME overcomes this limitation.
- ▪Ensemble averaging in GAME assigns weights based on data fit and function smoothness, enhancing reliability.
- ▪The study was published on the arXiv preprint server in February 2026.
Opening excerpt (first ~120 words) tap to expand
Space Astronomy Cosmology New AI algorithms are 95% better at showing how the universe changes over time A squad of new AI algorithms called GAME could help astrophysicists take a more accurate reading of the universe's changing behavior, a new study suggests. By Paul Sutter published 28 April 2026 in News When you purchase through links on our site, we may earn an affiliate commission. Here’s how it works. An illustration of a computer-rendered cosmos. A new suite of AI algorithms could help describe the nature of the universe with unprecedented accuracy, a new study claims.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Live Science.
Discussion
0 commentsMore from Live Science
