Meta Says Muse Spark Helped Solve 6 Major Math Problems
Meta says its Muse Spark AI models helped mathematicians make breakthroughs on six research problems while the company launched an open-source platform for Muse-compatible hardware.
Meta says its Muse Spark AI models have helped mathematicians solve six research problems spanning probability, differential equations, group theory, optimisation and other areas.
The company said the problems were open research questions rather than competition-style mathematics with predetermined answers. Researchers used Muse Spark 1.1 and 1.2 in Thinking Mode through the regular Meta AI interface, without custom modifications.
Among the reported breakthroughs, researchers used the model to identify a threshold involving random Gaussian points and ellipsoids, while another collaboration addressed wave collapse in a model inspired by laser physics. Muse Spark also helped researchers find a counterexample in group theory and contributed to work in optimisation, arithmetic physics and non-associative algebra.
Meta said mathematicians developed and checked the arguments, with a separate group reviewing the results.
The company is also expanding Muse beyond software with Muse Gadgets, an open-source ESP32 firmware and Linux SDK designed to help developers create hardware compatible with the AI assistant.
Meta also introduced Muse Home Link, a USB-C device that connects Muse to a home network and smart devices. The company said 5,000 units will be distributed to Muse subscribers.
AI solving maths and then getting its own hardware ecosystem? Apparently, even calculators are having an existential crisis now.
