Level 2: Gamers as an Untapped Resource
Leveraging millions of gamers to advance science.
Welcome to The Gameful Scientist! This newsletter explores the intersection of science and gaming. Enjoy!
What if instead of relying on a single research group’s efforts, we harnessed the collective intelligence of the masses? Two heads are better than one, but a million heads could be a game changer.
Often overlooked, gamers could hold the key to unlocking new frontiers in life science research. With their unique approach to problem-solving, willingness to dedicate time and attention to a task, and just sheer size, they have the potential to make a real contribution to science at a pace that traditional research can’t match.
In this article, we'll dive into the ways in which the gaming community can be leveraged for research, and how their collaboration with AI and the scientific community can lead to breakthrough discoveries.
Welcome to Level 2.
The Gamer’s Mindset
Gamers have a unique mindset that makes them great problem-solvers. They're used to being constantly engaged in virtual worlds, solving problems and overcoming obstacles, which shapes the way they approach and think about problems in real life.
For instance, gamers are accustomed to working within constraints and understanding the rules of a system. In Minecraft, players have to gather resources and build structures while sticking to the game's laws of physics. This translates well to research, where experiments are designed to test hypotheses within specific parameters.
Gamers are also great at thinking strategically and making decisions based on limited information, just like in the lab where experiments yield limited data and require interpretation. Most importantly, gamers are used to dealing with failure and uncertainty. As any scientist knows, you fail more often than you succeed. Much, much more. You embrace this failure to be successful.
Time is Our Greatest Resource
Time is precious, and yet we often take it for granted. Gamers, in particular, spend a significant amount of it in online, immersing themselves in addictive experiences. Take Fortnite for example, where players have spent over 10 million cumulative years playing the game.
On a personal level, some argue that gamers could find something "better" to do with their time. I don’t think that’s fair. Games provide a source of joy and community for many. But on a collective level, the opportunity cost of spending 10 million years playing a single video game is staggering. Imagine if even a small fraction of that time had been spent on social good. Or research.
Michael Retchin reframed this thought in terms of computational power. FLOPS, or floating point operations per second, is a measure that compares the performance of computers and estimates the processing power of the human brain. The fastest supercomputer in the world, Tianhe-2, in China, currently operates at around 54.902 petaFLOPS, while the human brain is estimated to have the capacity of 1 exaFLOPS, around 18 times more powerful. Obviously, not all this processing power is focused on a single task at any given time. But it’s an interesting thought. A significant cognitive capacity is devoted to games, and it's worth exploring ways to put it to better use.
The Scale of the Gaming Community
The gaming community is massive. There's an estimated 3 billion people playing games on computers, consoles, and phone devices this year, and the COVID-19 pandemic has only accelerated that growth as people turn to their controllers for escape and entertainment.
Citizen science projects like Zooniverse and discovery games like Foldit, Borderlands Science, and EVE Online's Project Discovery have shown the potential for these communities to make real-world contributions to science through data collection and problem-solving. The sheer scale of the gaming community allows for a level of participation in research that labs and institutes can't match. For example, Project Discovery is a minigame in the massively multiplayer online (MMO) game EVE Online. Players analyze data produced by flow cytometers and help scientists understand the human immune system's response to COVID-19. To date, 327,000 players have completed 1.37 million analysis tasks in-game, which has saved scientists 330.69 years worth of research.
Humans in an AI-Dominated World
These days, you can't open Twitter without seeing something about artificial intelligence (AI) or the fear of it replacing humans. And when it comes to the intersection of science and gaming, people often ask me why players are even necessary when AI will just be able to do everything.
But as AI advances, it's becoming more important for humans to play a role in its development. Take OpenAI's ChatGPT for example. It’s pretty good at understanding and responding to text. Really good. But its success is largely due to the human feedback it received during training.
But AI isn't just learning human language, it's also learning the languages of biology and chemistry. In life science research, this allows us to create completely new molecules that nature has never seen before. For example, NVIDIA recently partnered with a startup called Evozyne to create two brand new proteins, one for curing a congenital disease and another for consuming carbon dioxide to reduce global warming.
Dr. David Baker makes the claim that:
What we need are new proteins that can solve modern-day problems, like cancer and viral pandemics. We can’t wait for evolution. Now, we can design these proteins much faster, and with much higher success rates, and create much more sophisticated molecules that can help solve these problems.
Gamers can have a huge impact on advancing AI systems in life sciences by giving feedback through interactive simulations. Imagine playing a game that simulates different treatment options for a disease. This type of interactive simulation can provide valuable feedback for AI systems in personalized medicine, allowing them to better align with patient preferences, resulting in more effective treatments.
Or what about a synbio game? Games can provide valuable input for AI systems designing and optimizing synthetic biological systems such as bioprinting, genetic circuit design, and metabolic engineering. Kind of like the game Spore, but with modern tech and science. Maybe it can be integrated as a minigame in a future BioShock title!
By providing human input, expertise, and intuition that AI alone can't provide, gamers can play a crucial role in advancing AI systems and improving our understanding of complex biological systems.
Multidisciplinary Thinking in Research
Gone are the days of scientists working in isolation and making groundbreaking discoveries. In Reinventing Discovery: The New Era of Networked Science, Michael Nielsen writes:
The most important scientific discoveries are made at the interfaces between fields.Â
A recent study by Dr. Rhiju Das' group is a perfect example of this. They used a combination of the discovery game, Eterna, and a ML platform called Kaggle to solve the problem of RNA degradation. Players designed RNA sequences, and scientists measured their degradation rates. This data was then used as training data in a competition hosted on Kaggle. Over 1,600 teams participated for 3 weeks, and many of the entries outperformed the baseline models. To ensure that the models could generalize to new data and not just memorize the training data, the top two models were tested on a separate, unseen dataset. Both models performed better than other baseline models, showing the potential of using a combination of crowdsourcing and machine learning competitions to speed up scientific discoveries.
Data vs Expertise-Centric Discovery Games
Recently, I had a great chat with Josh Miller, a PhD student at Northeastern University's Khoury College of Computer Sciences and lead developer of Foldit. He explained to me how discovery games can be grouped by the way they source contributions from players.
Data-centric games, like Project Discovery, and Borderlands Science, rely on the gaming community to help with research by analyzing or collecting large amounts of data. Players don't need much training or expertise to make meaningful contributions. The scientific problem in these games is not necessarily hard, but there's a lot of data that needs to be processed by humans.
Expertise-centric games, like Foldit and Eterna, rely on the problem-solving skills of the gaming community to help with the research. These games challenge players to think critically and creatively to solve complex scientific problems. Through playing the game, players develop expertise and understanding of the problem, which allows them to generate new data and solutions that would be difficult or impossible for a small team of experts to discover on their own.
Motivation and Barriers in Discovery Games
Leonard Oppenheimer (LROppy) is a 68-year-old man in Minnesota who has spent 4 decades in the public sector and software industry. He's also the lead player of Eterna. He’s given me so much insight on the world of discovery games. We’ll explore several of these topics in later levels of The Gameful Scientist. He claims that:
Players play games like Eterna because they want to make a difference. In a world where large corporations have a major influence, they want to feel like their individual efforts can also make an impact.
This is an encouraging sentiment as the field works to design more effective discovery games. What’s stopping a million gamers participating in research projects? Certain challenges must be overcome to fully tap into this potential:
Time scale: Players often have to wait for results, which can clash with the immediate gratification of beating a boss or solving a puzzle that players are used to. To overcome this, rewards and mini-games can be implemented to increase engagement.
Evolution of discovery games: Discovery games get harder over time, which can impact player retention. For example, Eterna was built on the knowledge of RNA structure 12 years ago, but science has advanced since then, making the puzzles and understanding needed to solve current problems different than before. To overcome this, "power players" need to play a part in updating the game to reflect new scientific discoveries.
Instructional design: Discovery games have a steep learning curve, which can be overcome by including more advanced concepts in the tutorials and providing a gradual increase in difficulty.
UI/UX: Slow or outdated technology, and unintuitive gameplay can also be a barrier to player engagement. To overcome this, discovery games should use professional UI/UX designers, software developers, and game designers to create a more engaging and user-friendly experience.
Communication: Open communication between players and scientists, which can be overcome by providing clear instructions and resources and encouraging open communication.
There’s a delicate balance between making a game that’s good for science with one’s that actually fun to play. Make a AAA game that solves for this and you’ll unlock a massive community of players that can help you solve science’s toughest challenges.
I spend most of my time these days reading papers, chatting with scientists, game developers, gamers and investors to better understand this space. As I reflect on why I’m writing this article, I think about the influence gaming has on the younger generation. On culture. I imagine a future where instead of spending hours on Fortnite, young people are obsessed with solving real-world problems through games.
By creating opportunities for players to participate in real-world scientific research and problem-solving, discovery games can change the way the next generation views science and its impact on the world. It's not just about learning facts and figures, it's about being a part of something bigger. We can't expect the next generation to be excited about science if it's not presented in a way that resonates with them. By creating a community of Gameful Scientists, we can inspire new technologies, be an idea bed for founders, and show the power of collective intelligence to reimagine scientific discovery.
Join our community to discover the power of play! See you next week :)
Thanks for reading The Gameful Scientist! Special thanks to all the players and developers for the helpful chats. I’d love to hear your thoughts below.
Feel free to contact me here or chat with me on Twitter @ATrotmanGrant :)