Advanced Computing in the Age of AI | Thursday, October 6, 2022

You Gotta Know When to Fold ‘Em 

Not long ago, a group of gamers did in ten days what a world’s worth of molecular biologists had been unable to accomplish in more than a decade: collaborating online with a video game called Foldit, they identified the unique structure of a retroviral protease that lurks in a virus found in Rhesus monkeys.

Not long ago, a group of gamers did in ten days what a world’s worth of molecular biologists had been unable to accomplish in more than a decade: collaborating online with a video game called Foldit, they identified the unique structure of a retroviral protease that lurks in a virus found in Rhesus monkeys. The virus in question bears more than a small similarity to AIDS… and without the precise structure of this enzyme, scientists would have had essentially no chance of engineering a cure.

Protein folding is not something that computers are really good at yet, because computers aren’t great at spatial reasoning. Humans, who perceive the world spatially, can work with computers to accomplish feats that either alone wouldn’t have a chance at. Think of it as two-tiered collaboration: human beings working alongside other human beings, but also alongside computers. The world of human/computer interaction is a collaborative one, and growing more so with each passing day.

Two Views of the World

When we speak of high performance computing, we’re usually talking about applying massive processing power to solve complex problems in a comparatively short time. Whether simulating vehicle collisions or experimenting with new materials, it’s all basically the same. And in pretty much every instance, you still need people at the wheel. The computers can only do what they’re told; it takes users to make a computer into a worthy tool.

It’s important not to lose sight of the human involvement, and also important to remember that humans and computers look at things in very different ways. Foldit works by challenging people via a technique known as “gamification” – the application of play mechanics and game development philosophies to make the act of problem solving more engaging and enjoyable to the audience. That’s a fancy way of saying that video games might one day save the world… or at least a whole lot of lives.

Game designer Sid Meier once said, “A game is a series of interesting choices.” In the case of this protein-folding problem, gamers saw a challenging puzzle that happened to have real-world application. As Zoran Popovic, director of the Center for Game Science at the University of Washington noted, “Foldit shows that a game can turn novices into domain experts, capable of producing first-class scientific discoveries. We are currently applying the same approach to change the way math and science are taught in school.”

And well they should. Gamification is a new theory as far as homo sapiens is concerned, but education-through-play is a core part of all intelligent life’s development. By tying a game challenge to a powerful distributed computer network and encouraging collaboration between the players, we cracked a molecular code that had stumped us for more than ten years. By sticking their problem into Foldit and asking players to help out, researchers created a series of interesting choices by which an important solution could be reached. And since it was fun, gamers pounced right on it. “The game is not only an interesting intellectual challenge… but it also provides a unique society of players driven by both individual and team rivalry with an overall purpose of improving the game and the results achieved,” wrote one of the players.

How can these lessons be applied to a national digital manufacturing strategy? Well, first of all, we’re already talking about parallel collaboration between humans and computers. We’re dealing with engineering and manufacturing challenges that can be more easily solved through that collaboration. And of course one of the objectives of modeling and simulation is to solve in days or hours what would otherwise take months or years. It wouldn’t be too hard to gamify specific manufacturing challenges, either inside organizations that are on the cusp of adopting M&S tools or as a way to demonstrate to those still on the fence that this is a good idea.

Gamification and HPC

But it’s the collaborative aspect that may be most interesting. With the AIDS protein, scientists had a problem that they couldn’t solve the “normal” way, so they opened it up to the masses, providing a tool that would ensure any results would be scientifically applicable. The gamification provided by Foldit allowed players to collaborate and compete, and realize in-game benefits when they were successful.

Similarly, American manufacturers, faced with their challenges, and with access to modeling and simulation tools, could quite literally turn problem-solving into a collaborative game: distributed players with access to HPC collaborating (and competing) to realize new solutions made possible by the 21st century tools we all know are the Emerald City of smart manufacturing. The lesson of gamification, and Foldit, is that previously insurmountable problems can be shattered simply by offering the right tools and making the quest for a solution fun. Manufacturers who apply the same gamification philosophy to their own challenges may discover that they’re not just breaking down barriers, they’re having a great time doing it.

Add a Comment

EnterpriseAI