Not exactly this as far as official sources say, but the Karatsuba algorithm (long multplication in n^1.6) was the result of Kolmogorov (already a foundational

Not exactly this as far as official sources say, but the Karatsuba algorithm (long multplication in n^1.6) was the result of Kolmogorov (already a foundational figure, but only beginning his work in what would come to be called algorithmic complexity) starting a seminar with the goal of showing that long multiplication can’t be done in less than n^2 time (because nobody was able to do it faster in thousands of years) and a student (Karatsuba) coming in after a couple of meetings to prove him wrong (the catch being that the constant factors make the result not worth it for values of n people historically worked with by hand). (The algorithm boils down to the observation that (Xb + Y)(Zb + W) = XZ b^2 + [(X + Y)(Z + W) - XZ - YW] b + YW, for a total of three unique base-b products, plus what is nowadays a routine exercise to see that calculating these products recursively in base b/2 yields complexity n ^ log_2 3.)