The X factor

Our world is full of randomness. But in the programmed reality of computers, the truly random is both very rare and very sought-after. It can make software programs safer. It can help prediction models operate better. But how to achieve it? A 35-year-old associate professor at Cornell University has finally cracked the code, and has won the prestigious Godel Prize for doing so

- Kanika Sharma

Roll the dice, and the outcome could be anything between one and six. Such randomness fills our world.

Step into the binary reality of computers, though, and randomness becomes a rare resource, much sought after and largely unobtainable.

In the structured world of software programs, even computers tasked with generating a random result end up following a pattern of some kind. The closest they can come to true randomness is something called pseudo-randomness, where the patterns aren't easily visible and must be mined for.

Why does this matter?

Well, we don’t see it any longer, but there are a myriad ways in which software programs try to safeguard or hide the information they hold. Sometimes they do this via a PIN or OTP. Sometimes it is through the use of authentication or access tokens.

Asking a computer to be truly random when generating such safeguards is like asking a calculator to compose a poem. It simply isn’t programmed to do it.

In a world built on probability, could this gap ever be bridged? That is a question researchers have been asking since the late 1980s, from the Americans Gary Miller and Turing Award-winner Michael O Rabin to the Israelis Benny Chor and Oded Goldreich.

A 35-year-old associate professor at Cornell University has now arrived at something of an answer.

Theoretical computer scientist Eshan Chattopadhyay and his former doctoral supervisor David Zuckerman of University of Texas at Austin, have found a way to get computers to achieve something so close to true randomness as to be indistinguishable from it, by using two weak-random or pseudo-random strands of data.

Their efforts won them the prestigious Godel Prize, jointly awarded by the European Association for Theoretical Computer Science and the Association for Computing Machinery, in June.