(Of course) algorithms beat gamers at solving quantum control problems

Abstract

Dries Sels and Allan Grønlund have demonstrated numerical algorithms that outperform the players of our Quantum Moves game and they call the conclusions of our article Exploring the Quantum Speed Limit with Computer Games, Nature, 532, 210 (2016)] 'untenable'. Sels and Grønlund ignore that our article already emphasizes that the players do not solve the quantum optimization challenge to any desired accuracy. Our key finding is that player solutions can be employed either as seeds for numerical optimization algorithms or as guidance towards low-dimensional representations of the solutions.

A recent review of old numerical code revealed an error of implementation. An updated note will soon describe this and the implications for our work. We will also put up a description of recent much more thorough comparison of different algorithms and player solutions across a range of challenges with varying complexity to highlight strengths and weaknesses of player and algorithmic seeding