How can citizen scientists beat trained physicists?
Citizen scientists proved again that human intuition is better at solving certain tasks than complicated algorithms. It can be as simple as a game that even children can play. The Alice Challenge explored how citizens—without understanding the challenges in question—are able to contribute with valuable data. The article published in PNAS, “Remote optimization of an ultra-cold atoms experiment by experts and citizen scientists”, is a must-read if you want to get a deeper understanding of how your gameplay as a citizen scientist helping cutting edge research.
However, if the title is overwhelming or you hesitate to read it because you normally would not read a research paper, no worries. I will go through the main points of the paper in this blog post. If you are curious to find out more at the end of this post, I most certainly encourage you to read the paper yourself!
Can citizen scientists beat trained physicists?
In primary school, we learn to solve mathematical problems using a pen and a piece of paper. After having understood the fundamental rules of calculus, we move on and start to solve more complex problems using calculators. Across the field of natural science, we go even further and try to solve problems that are so difficult we need to use computers which can perform around 3 billion calculations per second. But even when we do so, we can run into problems which are too complex. It is nearly impossible to compute all the possible outcomes and see which one yields the best result.
An example of such a problem was what we introduced to the citizen scientists playing the Alice challenge in 2016. We gave the players live access to remotely control our experiment in the basement at Aarhus University. The task was to create the largest Bose-Einstein condensate possible, or you could say that their goal was to find the best solution to a problem they had no prior knowledge whatsoever.
A Bose-Einstein condensate is a weirdly behaving object, but for the purpose of this work, you can think of it as a cloud of atoms. Having a large number of atoms gives us a good starting condition for performing further experiments, so it's also an ideal state for the Alice Challenge. The citizen scientists were able to control two lasers and a magnetic field through an intuitive interface designed for this purpose. Whenever a solution was submitted, the player got back a score, indicating how good the solution was. Then, the player could further adjust the solution and send it back to the experiment.
You could imagine the difficulty of this challenge like asking a person who has never baked a cake before, to now bake a cake and make it better than a professional confectioner could. The task is laying out all the possible ingredients and finding the perfect balance between flour, sugar, vanilla, water, chocolate, eggs and whatever else that can be used to create the perfect cake. Every time a suggestion was submitted, the feedback would be a score telling whether the cake was good or bad. No more, no less.
Does it sound like an immense challenge to overcome? The same way, it is an immense task to ask people with no required training in physics to optimize the production of Bose-Einstein condensate in our experiment. However, players engaged in the challenge still found solutions that outperformed all the previous ones made in the lab by trained physicists.
Without any doubt, this result is amazing. The interesting question is: how is this possible? What is it about the human problem-solving technique that works so well? This is exactly one of the main topics of the paper. The Alice challenge was designed as a social science experiment aimed at extracting the collective behavior of the players while they were solving the task at hand. Hopefully, this acquired knowledge can help in the future to design new and more efficient ways to search for solutions across all scientific fields. Efficient ways to search for solutions, inspired by human problem-solving.
Are you interested to learn more about the Alice challenge and how human intuition can help find revolutionary solutions? Read the article on PNAS!
If you still don't feel like getting into science publications, listen to the conversation between Robert Heck and Carsten Bergenholtz, two of the scientists working on the research and publications about the Alice Challenge!