The next game-changing scientific discovery is as likely to come from an armchair amateur as it is from a researcher toiling away in the pristine chambers of a professional laboratory.
These days, thousands of citizen scientists are already readily answering the call of crowdsourcing sites to do the grunt work, brainwork and even funding of applied scientific research. And, instead of bristling at the prospect of noncredentialed, unschooled wannabes playing in their sandboxes, bona fide scientists are inviting them to jump in, sift through their data and even join in their games, some of which solve significant problems.
Examining data is the way the most unskilled aficionados are advancing science. Through the Zooniverse portal, run by the Citizen Science Alliance, citizen scientists comb through and scrutinize images of the Milky Way to search for stars that could host planet-forming disks, go through antebellum ship logs to recover Arctic weather observations and examine fathoms of ocean to ascertain the size and floating direction of plankton. For more than 1 million citizen scientists logging in, it's an unskilled, unpaid labor of love.
In those cases, the images are supplied; other crowdsourcing sites rely on amateurs to also provide the pictures. And it turns out that theirs are often as reliable as those taken by professionals. Citizen scientists recently passed their first published litmus test in conservation science literature: There was "virtually no difference" between photographs identifying whale sharks via markings submitted by citizen scientists and those of marine biologists.
"These collaborating 'non-researchers' are able to provide the same quality photos as bona fide whale shark researchers," Tim K. Davies, the study's author, a marine biologist at the Imperial College London, U.K., tells Newsweek. Even if it was saddled with a greater margin of error, citizen science would still be "absolutely" worth it, he says. "Crowdsourcing is free, and long-term field research is not."
Sharing the expense and workload of long-term field research with willing neophytes is a godsend to astronomers, geologists, marine biologists and climate scientists—all of whom are inundated with the dizzying amounts of data churned out by satellite images. It's also a huge boon to geneticists, who are often overwhelmed with the amount of unique pieces of data that result from large-scale DNA sequencing. They're itching to crunch it all, but simply can't on their own.
Kevin Campbell, a University of Manitoba professor of environmental and evolutionary physiology who specializes in analyzing the DNA of extinct animals, says he would welcome the assistance of citizen scientists to decipher the ancient genetic material he's now inundated with, thanks to a thawing Arctic.
"We now can collect so much data in so little time—hundreds of millions of DNA sequences that we have to go through by hand because of ancient DNA damage," Campbell tells Newsweek. "We have not near enough time to analyze it. With crowdsourcing, you can get the public to do the work for you by making a game out of it."
Indeed, crowdsourcing sites that turn even the most serious research into a "Can you solve this puzzle?" project are often irresistible to citizen scientists, who hope to test their ingenuity against that of their online peers. On Kaggle, users can compete to solve real puzzles posed by industries and universities, and a leaderboard keeps track of which competitors are ahead in the "game." The results, though, are serious. For example, Kaggle users recently worked on algorithms to predict the health of patients with HIV by finding markers in their HIV sequences measured by viral load and CD4 counts. The winner was a very unlikely scientist: Former literature major and college dropout Chris Raimondi of Baltimore beat out 106 other teams by creating an algorithm that predicted changes in HIV infection severity with 78 percent accuracy. He's on the verge of publishing a paper on his work.
"Our site democratizes the scientific process," Anthony Goldbloom, the founder and chief executive of Kaggle, tells Newsweek. (Although this might be semantics, Goldbloom considers his site more of a hosting site than a crowdsourcing one.) "It gives data scientists, and other people who don't happen to have a Ph.D. working full time at university, a chance to solve really challenging and important problems that benefit society." Indeed, about two thirds of the 150,000 or so statisticians who make up the Kaggle community, Goldbloom estimates, are not academics.
One competition on Kaggle, sponsored by the William and Flora Hewlett Foundation, rated the essay-scoring software of the eight largest testing vendors in the U.S. to see if they could match expert hand graders for accuracy. (They did.) The competition also challenged contestants to come up with their own software. Carnegie Mellon University grad student Elijah Mayfield did so well with his scoring engine, LightSide, that he formed his own company and has schools interested in a pilot project. Kaggle has been so impressed by its contestants that it's plucked the ones from the top of its leaderboard to work in its own company.
The brainpower of the citizen scientist is not lost on the federal government, either. NASA has many popular crowdsourcing projects, including an asteroid detection initiative launched in March. Another, launched in January, asks citizen scientists to search Hubble Space Telescope images for star clusters. The National Science Foundation (NSF) has made crowdsourcing part of its 2014-2018 strategic plan and helps fund such crowdsourcing programs as the puzzle-solving site foldit, which enjoins visitors to "Solve Puzzles for Science."
For example: The chains of amino acids that comprise a protein fold into a complex, three-dimensional, often globular arrangement. Knowledge of the folded structure would allow therapies to target them perfectly. Being able to predict how a protein looks when it's folded allows for the development of synthetic proteins, which can then be used to interlock with cancer cells or the HIV virus long enough to kill them. But the number of different ways even a small protein can fold is virtually infinite. According to the website, "Figuring out which of the many, many possible structures is the best one is regarded as one of the hardest problems in biology today."
And yet the citizen scientists do figure it out. They were able to ascertain the structure of a monkey virus similar to HIV in 10 days, a problem that had stymied researchers, and computers, for years. Artificial intelligence couldn't do the trick, because it lacks the ability to figure out complex 3-D arrangements (just as it has yet to match human aptitude for face recognition).
The crowd is also beginning to move scientists toward those projects that interest it the most. On the for-profit site Petridish, the public pledges funds to the science research projects they find most appealing. Although there's nothing to stop visitors from pledging to more than one project, there is a competitive feel to the site, where the most compelling projects (often those with the best presentation) meet their funding goals by deadline. As with Kickstarter, projects that don't meet their goal get nothing, and pledgers are usually rewarded with prizes such as trinkets from the field, lecturers or dinners with the scientists and even acknowledgements in professional journals.
The site has a decidedly glossy, travel brochure-like feel, with beckoning photos of the animals that are the object of field projects, and blog-like presentations from likable scientists. It's not hard to see the handiwork of one of Petridish's two founders, Ilia Papas, who has done e-commerce work for Wal-Mart and Macy's.
"It becomes a popularity contest," Campbell says in reference to science research sites that involve voting. "People selecting projects on 'cute' or 'iconic' species, or those that are 'personal'—a cancer survivor voting for a cancer study solely because they themselves have or had the disease." The geneticist points out that as well-meaning as the public may be, the fact that the research projects are proposed out of context may sway the voting.
"Most of the public doing the 'selecting' likely do not have the relevant background to assess the scientific merits of the research itself," Campbell says. "They may be unable to determine potential pitfalls or shortcomings of the proposal." The public may not recognize when a proposed project is basically just a duplicate of previously published research, or whether it's even feasible.
Kevin Crowston, a program director in the NSF's Computer and Information Science Program, agrees. "Public voting might not be the best way to identify important basic research," Crowston tells Newsweek. "And to substitute the evaluation of the crowd for peer review to select projects—I don't see it working, though it could be an interesting adjunct."
Crowston downplays the role the citizen scientist will play in the future of scientific research. "I don't think the crowdsourcing is likely to drive the direction of science, since the impetus from the projects comes from scientists rather than the crowd for most projects," Crowston says.
But others believe that sites like Zooniverse are laying the groundwork to push their contributors—previously amateurs—into conducting their own legitimate research.
"There's an interesting emergent effect of engaging citizens in scientific research," explains David Weinberger, a philosopher at Harvard's Berkman Center for Internet & Society and author of Too Big to Know, a book about networked knowledge. "Citizens may start out being treated as sensors, reporting on local conditions, or as very simple computers." But then their work naturally becomes more nuanced and complex.
Weinberger recalls, for example, that when Zooniverse asked citizen scientists to classify images of galaxies as circles or spirals as part of its GalaxyZoo project, they began reporting odd formations. "Because the site enabled them to talk with one another, those anomalies got noticed, and so-called 'green peas' became objects of scientific study," Weinberger says. "But because we're a social species, we won't stay as sensors or simple computers. Given a chance, we'll connect, and that network of connected citizens may advance science beyond the original expectations."