Researchers Use Gemini 4 and Multi-Agent Systems to Solve Decades-Old Scientific Open Problems
Researchers successfully deployed Gemini 4 and multi-agent harnesses like /teamwork and Stellar Colosseum to independently solve long-standing open problems across computer science, quantum computing, and mathematics.
🤖AI ModelsPhoto: Hacker News Front Page
Key points
- Gemini 4 and multi-agent harnesses solved decades-old open problems in math and computer science.
- Multi-agent systems independently proved and Lean-verified near-optimal sample bounds for adaptive data analysis.
- New deterministic algorithms improved matrix multiplication, integer multiplication, and optimization runtimes.
A team of researchers has demonstrated significant breakthroughs by utilizing advanced AI models and multi-agent frameworks to tackle complex, unresolved problems in engineering and science. By posing open challenges directly to Gemini 4 alongside multi-agent systems such as /teamwork and Stellar Colosseum, the investigators received end-to-end algorithms, mathematical proofs, and formalizations verified in Lean 4.
Among the notable achievements, a multi-agent harness running on Gemini 4 Argon proved near-optimal sample bounds for adaptive data analysis, effectively resolving a statistical query gap that had persisted since 2014. In quantum computing, a newly developed construction settled a decade-old question regarding n-bit Boolean functions, while other efforts yielded deterministic algorithms for Boolean matrix multiplication and major progress on Tarjan's 1985 Deque Conjecture.
The projects also advanced optimization, randomized linear algebra, and data structures. For instance, campaigns using these AI systems produced a deterministic algorithm for online inverse linear optimization that reduces runtime from exponential to polynomial time, alongside a new $O(n \log n)$ integer multiplication algorithm whose C implementation matches the GNU Multiple Precision Arithmetic Library bit-for-bit up to a billion bits.
What to watch next: Look for further adoption of multi-agent AI harnesses in academic research as scientists increasingly rely on automated verification tools like Lean 4 to validate machine-generated proofs.
Source Hacker News Front Page — research.google/blog/advancing-science-engineering-with-antigravitys-te… ↗
Read next
🚀Startups & FundingHarvard Study Reveals US Hyperscale Data Centers Consume Power Rivaling Entire States
A new Harvard study reveals that America's 403 hyperscale data centers consumed between 68 and 99 terawatt-hours of electricity during the 2024-25 study period, rivaling …
🚀Startups & FundingCalifornia Files Lawsuit Against OpenAI Over Autonomous AI Incidents
California Attorney General Rob Bonta has filed a lawsuit against OpenAI, alleging the company failed to control its autonomous AI agents and misled the public about safe…
🧠AI AnalysisPedestal AI Secures Seed Funding to Automate Logistics With Intelligent Agents
Pedestal AI has launched publicly with $4.7 million in seed backing to deploy agentic artificial intelligence aimed at streamlining logistics and supply chain operations.
🏢Big TechUnited Launch Alliance Investigates Vulcan Rocket Failure
United Launch Alliance is investigating a mission failure of its Vulcan rocket, which suffered from low booster performance shortly after liftoff. The incident raises que…
🌐Work & SocietyUS Southwest Hydropower Output Hits Record Lows Amid Persistent Drought
Hydropower generation in the US Southwest dropped 28% below the 10-year average this summer as record-low water levels in the Colorado River Basin threaten regional energ…
🏢Big TechYouTube Updates View Metrics to Include TV Co-Viewing Estimates
YouTube is incorporating estimated co-viewing data into its public view counts to better reflect group consumption on television sets.