Researchers at Google DeepMind have developed FunSearch, an AI system that successfully discovered new solutions to complex mathematical problems like the cap set problem. By combining a large language model with an automated evaluator, the system generates verifiable computer code that outperforms existing human-designed algorithms.
TLDR: Google DeepMind’s FunSearch has achieved a milestone by using large language models to discover new mathematical solutions. By pairing an LLM with a rigorous evaluation system, it solved the cap set problem and optimized bin-packing algorithms, providing human-readable code that advances both theoretical mathematics and practical computer science.
Researchers at Google DeepMind have unveiled a new artificial intelligence system called FunSearch, which has successfully discovered novel solutions to longstanding problems in mathematics and computer science. The system represents a significant shift in how large language models are utilized for scientific inquiry. Rather than acting as a simple conversational interface, FunSearch operates as a specialized search engine that explores function space to find programs that solve specific puzzles. This methodology allows the AI to move beyond existing human knowledge to generate entirely new insights.
The architecture of FunSearch involves a collaborative loop between a pre-trained large language model and an automated evaluator. The system is tasked with writing computer code in the Python language to solve a given problem. These code snippets are then passed to the evaluator, which runs the code to verify its correctness and measure its performance. Only the most successful programs are kept and fed back into the model for further refinement, creating an evolutionary process of improvement that filters out incorrect or inefficient logic.
One of the primary achievements of FunSearch involves the cap set problem, a central challenge in the field of combinatorics. This problem asks for the largest possible set of points in a high-dimensional grid where no three points form a straight line. For decades, mathematicians have struggled to find the exact sizes of these sets as dimensions increase. FunSearch discovered a cap set in an eight-dimensional grid that was larger than any previously known, marking the first time an artificial intelligence has been used to make a new discovery in a major open problem in mathematics.
Beyond theoretical mathematics, the system was applied to the bin packing problem, which is a staple of logistics and computer science. This problem involves finding the most efficient way to pack items of different sizes into the fewest number of containers. FunSearch generated new algorithms for this task that outperformed those designed by human experts over several decades. Because the system outputs code rather than just a numerical answer, the resulting algorithms can be easily inspected, understood, and deployed in real-world industrial settings.
A key advantage of the FunSearch approach is its transparency. Traditional deep learning models often function as black boxes, providing results without explaining the underlying logic. In contrast, FunSearch produces programs that humans can read and analyze. This allows researchers to understand why a particular solution works, potentially leading to further insights that the AI itself did not explicitly state. The system also avoids the common issue of hallucination because every output is strictly verified by the automated evaluator before being accepted.
The success of FunSearch suggests that artificial intelligence is moving toward a role as a proactive collaborator in the scientific process. By automating the search for new ideas and verifying them in real-time, researchers can bypass the trial-and-error phases of mathematical discovery. The DeepMind team intends to apply this methodology to a wider range of scientific domains, including materials science and drug discovery, where finding optimal configurations is a recurring challenge.
As AI models continue to scale, the integration of rigorous evaluation frameworks will be essential for ensuring the reliability of their outputs. The development of FunSearch demonstrates that when paired with the right constraints, generative AI can move beyond mimicking human language to uncovering fundamental truths about the mathematical world. Future research will likely focus on expanding the types of programming languages the system can use and improving the efficiency of the evolutionary loop.
Mason Reed serves as a Staff Writer for Just Right News, where he spearheads the Future Frontiers & Special Projects desk. In an era defined by rapid technological shifts and evolving social landscapes, Mason provides a steady, principled voice, examining the innovations of tomorrow through the lens of traditional American values. His work is most prominently featured in his signature series, “The Next Horizon,” where he explores the intersection of emerging technology, national sovereignty, and the preservation of individual liberty.
A native of San Diego, California, Mason’s worldview was shaped by the unique culture of his hometown. Growing up in a region defined by its strong military presence and its history of maritime industry, he developed a deep-seated respect for the institutions that provide national stability and the entrepreneurial spirit that drives the American economy. This upbringing instilled in him a belief that true progress is not found in discarding the past, but in building upon a foundation of proven principles. His reporting often reflects this San Diego influence, emphasizing the importance of a robust national defense and the necessity of maintaining a competitive edge in the global marketplace.
Now based in San Francisco, Mason operates from the heart of the world’s technological engine. Living and working in the Bay Area provides him with a front-row seat to the advancements—and the ideological challenges—emanating from Silicon Valley. While many in the region embrace a “move fast and break things” mentality, Mason’s reporting serves as a vital counterweight. He offers Just Right News readers a “boots on the ground” perspective, documenting how radical local policies and the concentration of tech power impact the everyday lives of citizens. His proximity to the industry allows him to cut through the marketing jargon of big tech to uncover the real-world implications for privacy, free speech, and the nuclear family.
In his “Future Frontiers” beat, Mason tackles complex subjects ranging from the ethics of artificial intelligence to the burgeoning private space race. He approaches these topics with a healthy skepticism toward centralized bureaucracy, championing instead the decentralized innovations that empower individuals. Through “The Next Horizon,” he highlights the pioneers and thinkers who are working to ensure that the future remains a place where human dignity and constitutional rights are protected. He believes that the rapid pace of change requires more than just technical expertise; it requires a moral compass rooted in the Western tradition.
Throughout his tenure at Just Right News, Mason has remained committed to the idea that the future is something to be shaped, not merely accepted. His writing is characterized by a rigorous defense of American exceptionalism and a belief that the country’s best days lie ahead, provided it remains true to its founding ideals. Whether he is investigating the impact of automation on the American workforce or profiling the next generation of aerospace engineers, Mason Reed ensures that his readers are equipped with the insights they need to navigate a changing world with confidence and clarity.