-
💡Matrix multiplication is a complex mathematical operation that appears in various fields, and finding faster methods for it can solve larger computational problems that were previously considered too big to be computable in a reasonable time.
-
🧮The standard algorithm for multiplying matrices takes N-cubed steps, making it inefficient for larger matrices, but Volker Strassen discovered a faster algorithm that only requires seven multiplication steps.
-
🚀A new algorithm, revealed in October 2022, surpassed Strassen's algorithm for multiplying two four by four matrices with elements of zero or one, demonstrating the potential for even faster matrix multiplication by breaking them into four by four matrices instead of two by two's.
-
🤯DeepMind's AI algorithm, AlphaTensor, was able to tackle the challenging problem of matrix multiplication, which even for small cases, didn't have an optimal solution.
-
💡The use of tensors in matrix multiplication algorithms can potentially lead to more efficient calculations and improved performance.
-
🧠Formulating the problem of finding the most efficient matrix multiplication algorithm as a clearly defined computer task allows for the application of search and machine learning techniques to explore the large search space.
-
🚀AlphaTensor's algorithm broke a 50-year record by using only 47 multiplications to multiply two four by four matrices in modulo-2, surpassing the previous record of 49 multiplications by Strassen's algorithm.
-
🤝Rather than making people irrelevant, advancements in AI empower individuals to accomplish more.Researchers at Google research lab DeepMind trained an AI system called AlphaTensor to find new, faster algorithms to tackle an age-old math problem: matrix multiplication. Advances in matrix multiplication could lead to breakthroughs in physics, engineering and computer science. AlphaTensor quickly rediscovered - and surpassed, for some cases - the reigning algorithm discovered by German mathematician Volker Strassen in 1969. However, mathematicians soon took inspiration from the results of the game-playing neural network to make advances of their own. Read the full article at Quanta Magazine: https://www.quantamagazine.org/ai-rev...