The rise of human computation

In a detail too often lost to history, the first “computers” in businesses and organizations weren’t digital devices at all. They were human beings, usually women, hired to perform complex calculations. These “computers,” as they were known, put the first men on the moon.

In time, human computers were replaced by the computing tools we know today — tools that have exploded in ubiquity and power. These technologies are poised to overthrow everything we think we know about how people work and how companies make decisions. As self-driving cars take to the roads, self-driving organizations are also revving up their engines.

What will be the role of human beings in this new economy? Will we just be passengers, along for the digitally-driven ride? If we go about it thoughtfully, we can be so much more — but that may mean flipping the script and rethinking the human/computer dyad.

Making the most of what makes us human

In many ways, computers drive organizations today. And our reliance on computers is only growing. Enterprise systems use technology to collect, manage, and analyze data. They also rely on technological solutions to link smart devices, facilitate transactions, connect people, and coordinate resource exchanges.

Yet despite these advances, the human mind is (for now at least) far more capable than a computer. And so a new field is emerging that explicitly combines the contributions of people and machines: human computation.

The field goes by many names: human-based computation (HBC), human-assisted computation, ubiquitous human computing and distributed thinking. By any name, this emerging field comprises a multidisciplinary community of academics, visionaries, private industry researchers, and federal program officers. Together they’re exploring the transformative potential of “directly employing human cognition within larger computational systems.”

Why human computation? Distributed systems combine the strengths of human cognition and electronic computation. Together they can do what neither can accomplish alone. Human computation is even a critical part of artificial intelligence. Uniquely human work creates the data sets that feed cognitive computing and deep learning.

Human computation systems have evolved in three key ways:

• Crowdsourcing: Large tasks are broken down into microtasks that are distributed to a large number of humans via a user-friendly interface; data is later aggregated for further processing.
• Complex workflows: Data is funneled to crowd workers in specific roles, with workers at each level building on information provided by previous workers.
• Problem-solving ecosystems: Researchers use human and machine-based contributions inform models of complex systems and global challenges.

You’re already part of the machine

Distributed thinking may sound novel, but odds are you’ve already acted as part of the model. Ever used reCAPTCHA? Duolingo? How about Wikipedia?

Since its inception in 2001, Wikipedia has become the internet’s key resource for all kinds of basic information — and a perfect example of the human computation model. Digital assistants like Alexa, Siri, and Google Assistant pull answers from more than 41 million human-created articles. Some 70,000 active contributors — mostly anonymous and unpaid human writers and editors — work together in nearly 300 languages to create the free content encyclopedia.

The seminal example of human computation was created by Luis von Ahn, the legendary digital pioneer who coined the term itself. reCAPTCHA is a web widget that, on the surface, keeps websites safe from bot attacks. Every day, people solve hundreds of millions of CAPTCHAs (Completely Automated Public Turing tests to tell Computers and Humans Apart) to prove they are indeed human.

But the effort that goes into verifying one’s humanity doesn’t go to waste. People prove their human-ness by completing tasks that are difficult for software alone to perform. Hundreds of millions of microtasks have been leveraged to annotate images, build a machine learning dataset, improve maps, and solve hard AI problems. Human computation in reCAPTCHAs have fed digitization of Google Books and over 13 million articles from The New York Times archives.

Von Ahn also founded Duolingo, an app-based language learning program. As of March 2016, Duolingo had reached 110 million users worldwide. Like reCAPTCHA, the Duolingo software has benefits on both sides of the equation: it offers free language lessons while simultaneously acting as a document translation service. Because machines struggle to learn and process natural language, human computation has been an essential element.

Beyond data: solving problems together

Human computation can offer so much more than collections of human-provided data. Fold.it, for example, is one of the field’s most successful applications. The project is well-known in the citizen science arena, where members of the public contribute to scientific research. Citizen science brings the power of the crowd to address questions that are otherwise too big for the lab.

The goal of Fold.it was to understand protein folding in molecular biology. Contributors to the Fold.it project used a gaming platform to fold virtual proteins as efficiently as possible. Human competitors brought their spatial reasoning abilities — another struggle for computers — to the problem. In three short weeks they had contributed to the fight against the AIDS virus by solving a protein structure problem that had plagued researchers (and their powerful computers) for 15 years.

Human computation can also help address so-called “wicked problems,” whose complexity makes them resistant to easy resolution — problems like climate change and global conflict. Although microtasking has proven effective, it won’t be enough to conquer wicked problems: we’ll need to find ways to get even more participants and make the most of their contributions.

Human potential in a digital age

Leaders in the field have argued that there needs to be a new national initiative in human computation, with the creation of a national center devoted to this emerging science. According to Janis Dickinson, professor and director of citizen science at the Cornell Lab of Ornithology:

“By providing the right kinds of information about our cooperative efforts and where we stand as cooperators, human computation systems can provide unprecedented support for people to help work on large problems that require large-scale human effort to solve.”

Public participation in human computation can create citizen-informed R&D, drive product value and contribute to brand meaning. The discoveries we make through these collective, public undertakings represent yet another layer of human potential in increasingly software-driven organizations.