By Christina Arthur
How can business teams boost their collective intelligence? That’s a question Dr. Lynn Metcalf of Cal Poly’s Orfalea College of Business and Dr. David Askay of Cal Poly’s Communication Department set out to answer when they collaborated with Dr. Louis Rosenberg, founder of Unanimous AI, a Silicon Valley-based technology firm, in a new article about Artificial Swarm Intelligence.
The 26-page paper, which was published in California Management Review of Berkeley’s Haas School of Business, is titled “Keeping Humans in the Loop: Pooling Knowledge through Artificial Swarm Intelligence to Improve Business Decision Making.” At its core, it introduces Artificial Swarm Intelligence as a collaboration technology that can enhance traditional AI by harnessing collective human brain power.
It also goes into depth about how business teams can be benefitted by Swarm AI technology, boosting their intelligence above traditional methods to create efficient tools for a broad range of applications, from forecasting financial assets, to predicting sales through optimizing group input.
“The Swarm AI tool is a way to integrate both humans and machine-learning to make a decision together. [The goal] is to keep humans in the loop and to have some sort of influence over the results.”
As Askay explains, artificial intelligence usually operates on pre-existing data received from machine-learning and the patterns it identifies, a process that typically removes human input from the equation. “There are some concerns about removing people from this process,” he says, “so the Swarm AI tool is a way to integrate both humans and machine-learning to make a decision together. [The goal] is to keep humans in the loop and to have some sort of influence over the results.”
He further explains this human element is an important component, since the mind possesses intuition and implicit knowledge that’s hard to articulate. Sometimes, for instance, even when weighing complex decisions, a person will operate on a hunch, or an instinct that they are making the right decision, but they aren’t sure why. In an ideal scenario, these conclusions are based on experience and accumulated knowledge. Swarm AI allows for those instincts to be heard, without the human messiness of having to explain the reasoning behind each decision.
Metcalf adds this tool is different from other AI because it engages humans in a collaborative process of decision-making. In some cases, she says, it has demonstrated that humans, when we pool our intelligence, are capable of a greater wisdom or collective intelligence than machine learning. She also says that at first, the concept of Artificial Swarm Intelligence was envisioned as a type of fun, collaborative tool that networked groups could use inside chat rooms to quickly reach decisions and converge on group forecasts, with teams of Orfalea College students identifying outside-the-box applications for the tech.
“Then Unanimous AI got some really interesting results on predictions for events like the Oscars and the Kentucky Derby Superfecta,” she says, “and it started to take off as a rigorous decision-making tool that can be used for forecasting in business situations.”
She also says the Swarm AI tool can predict a wide range of outcomes because it’s not topic-specific—weighing in on the results of the World Series to predict which teams will triumph, as well as business-use cases such as which sweaters will sell, or which movies will have the best results at the box office.
“Unanimous AI got some really interesting results on predictions for events like the Oscars and the Kentucky Derby Superfecta and it started to take off as a rigorous decision-making tool that can be used for forecasting in business situations.”
According to Metcalf, Swarm AI tools are currently being used in pockets of industries, particularly in financial organizations for predicting markets. She says she firmly believes the technology has the potential and the ability to be widely used and that the future of Swarm AI depends on how deeply businesses integrate the tool into their daily activities, which of course depends on teams experiencing real-world outcomes that validate performance gains.
“As an outsider, it’s easy to become skeptical of it,” Askay says, “which I think is the biggest hurdle. It’s just like when any new product comes along. It takes some time to adjust.”
Metcalf and Askay both express their hope for the future of Artificial Swarm Intelligence and how its integration could address certain concerns about artificial intelligence and its rapid advancement. “There’s some sense of urgency about keeping humans in the loop on artificial intelligence,” says Metcalf, “so that humans aren’t relegated to executing the will of machines.”
To read more about the Swarm AI tool and its case studies, click here.