A review of recent research on recursive self-improvement (RSI) concludes that while large language models have made remarkable advances in scientific tasks, every demonstrated success relied on an ex…
Google and DoIt used AlphaEvolve, Google Cloud's autonomous evolutionary code-optimization tool, to optimize production Swift code in a live macOS streaming app, uncovering performance headroom that m…
Google researchers and academic collaborators described Dream RSI, a method that lets an AI system improve its own strategy for guiding scientific discovery by simulating past research history rather …
Google researchers, with collaborators from Google DeepMind, the University of Maryland, and the University of Virginia, published a paper describing Dream-RSI, a system that recursively improves an A…
An AI agent built on Google's AlphaEvolve pattern beat 10 listed Packomania circle-packing records overnight for a total model spend of $27.72, with improvements ranging from 2.5% to 5.4% above prior …
A multi-agent system of large language models autonomously solved five major open problems in mathematics, including new finite-field Kakeya sets and improved bounds on Erdős's minimum-overlap problem…
Jerry Tworek, former OpenAI VP and current CEO of Core Automation, estimates human researchers have at least two years left as a meaningful part of AI research discovery, citing the exhausting pace an…
In the Station, an open-world multi-agent environment where AI agents from different model families pursue shared research goals without central coordination, agents achieved novel mathematical result…
Google DeepMind's AlphaEvolve, an evolutionary coding agent built on its Gemini architecture, has refined the matrix multiplication exponent omega from 2.371339 to 2.371177, marking the tightest bound…
Google DeepMind, with academic collaborators and its AlphaEvolve coding agent, has lowered the theoretical upper bound for matrix multiplication exponent ω to less than 2.371177, improving on the prev…
Researchers using a machine-learning-guided optimization algorithm called AlphaEvolve have improved the upper bound on the matrix multiplication exponent to ω < 2.371177, surpassing the previous best …
A new formulation for designing AI 'harnesses' classifies task topology by feedback density, search geometry, horizon, evaluation cost, and constraints, then chooses how explicitly search should be co…
Public data on exploited software vulnerabilities and solved open math problems shows a sharp acceleration in discoveries in 2026, but aggregate algorithmic optimization records show no clear change i…
Researchers introduced ε-MemEvo, a framework for cross-task knowledge transfer in LLM program evolution, which stores prior experience as task-agnostic tactic memories and uses an adaptive injection g…
Ryanair has signed a five-year agreement with Google Cloud to deploy Google Workspace, Gemini Enterprise agentic AI, DeepMind's AlphaEvolve, and WeatherNext across its 35,000-strong workforce, weeks a…
Ryanair announced a five-year partnership with Google Cloud on August 12, 2026, to integrate Google Cloud services and Google Workspace across its 35,000-employee network, adopting a dual-cloud strate…
Tom Zahavy, a Google DeepMind researcher and co-author of AlphaProof, argues in an ICML 2026 position paper that large language models have mechanized induction and deduction but lack a mechanism for …
Phylo introduced Biomni x TusoAI, a fully autonomous and code-free tool for biology method development that builds state-of-the-art models for genetic perturbation prediction and enhancer-gene linking…
A new framework called MetaEvolve, developed by researchers and detailed in a paper on arXiv, teaches large language models to self-evolve through reinforcement learning, achieving a 10.01% absolute i…
Google has released AlphaEvolve, an AI coding agent developed by Google DeepMind, on the Gemini Enterprise Agent Platform. The tool uses large language models to automatically generate and improve alg…