Humans Triumph Over Machines In Protein Folding Game FoldIt

Illustration for article titled Humans Triumph Over Machines In Protein Folding Game FoldIt

Two years ago students and instructors from the University of Washington applied human intuition to the problem of protein folding with the game FoldIt. Since then the game has proven that even under-educated players can out-fold the most powerful computers.


Proteins are essential biological macromolecules that have an active role in just about every process that goes on inside a cell, from maintaining cell structure to managing immune reactions. Made up of a chain of amino acids folded into a globular form, understanding the structure of folded proteins could be the key to countless medical advances.

The problem is, the act of folding proteins into the correct shape is a job that would take all the computers in the world centuries to solve, which is why University of Washington professor of biochemistry Zoran Popovic and his colleagues decided to apply a more human solution to the problem. That solution is the online protein folding game, FoldIt.

Now the results are in, published in the journal Nature, demonstrating that human intuition can indeed compete with supercomputers, especially when the problems involved require radical moves, risks and long-term vision.

Comparing the performance of humans against the performance of the Rosetta distributed computing protocol for folding proteins, humans excelled in solving problems that involved making intuitive leaps or shifts in strategy. While the computer program must follow a strict set of guidelines and procedures, humans are free to change things up.

Players from all walks of life and levels of education have flocked to the title, performing solo or in teams for spots on FoldIt's leaderboards. The competitive nature of the game is one of the keys to its success, and a key factor in drawing together the more than more than 57,000 FoldIt players mentioned in the author list for the published paper.


What's even more impressive about FoldIt's success is the wide variety of players who have participated. The following chart shows the demographic makeup of the top 50 FoldIt players in two categories, Soloists and Evolvers.

Illustration for article titled Humans Triumph Over Machines In Protein Folding Game FoldIt

While there is a strong tendency for the top players to be American males, occupation and education levels are all over the place. It doesn't matter if you didn't finish high school and don't have a job, you can still lend your intuition to the protein folding problem.

"It's a new kind of collective intelligence, as opposed to individual intelligence, that we want to study,"Popović said. "We're opening eyes in terms of how people think about human intelligence and group intelligence, and what the possibilities are when you get huge numbers of people together to solve a very hard problem."


When we reported on FoldIt in 2008, they had just moved on from testing with solutions to known proteins to problems with unknown solutions. Now the players will focus on creating entirely new protein structures. The first attempt at synthesizing a player-created protein took place last year at the lab of co-developer and University of Washington biochemistry professor David Baker, and while that attempt failed, Baker has high hopes for the future.

"I think that design problems are an area where human computing has huge potential," Baker said. "People are good at building things, so I'm expecting that people will be very good at building proteins for different purposes. That's where I'm expecting really great things from FoldIt."


One day FoldIt players could design proteins that battle HIV and other incurable diseases, but for right now they are content simply playing to win.

For more on the FoldIt progress report, check out the University of Washington press release. The full report can be purchased online at the Nature journal website.


Thanks to Nick for pointing us to the report!



I doubt computers will ever be able to outperform humans in the domain of intuition and creativity.

They MAY get close, but the programming language (correct me if I'm incorrect) has its limitations. Even then, we'll most likely be extinct before that happens.