The Human Side of Artificial Intelligence

The world is awash in data. There is more information now than at any time in human history and we’re creating more of it faster than ever before. Powerful artificial intelligence-driven algorithms analyze that data not just to understand what’s already happened, but to also predict what’s likely to happen next—with transformative implications for human health, prosperity and economic growth.

AI is already “helping companies manage risk and ensure strong balance sheets, liquidity and shareholder value,” says Cathy Bessant, Chief Operations and Technology Officer at Bank of America. And at the same time, “We’re getting better at meeting customer needs in real time.”1
   
Yet as with any transformative technology, AI presents its own risks, including the potential for unintended consequences. By focusing too heavily on how to use it, businesses and governments risk overlooking a subject of equal or greater importance— how do they ensure that it is used wisely?

Towards responsible AI

The answer won’t be found in an algorithm, it can only be provided by people and requires reframing how we think about artificial intelligence. According to Bessant, it’s critical to understand that AI is a subset of human intelligence, not a replacement for it. “Technology is not the destination,” she says, “and AI will not create better decisions on its own.”2

As a product of humans, AI is subject to the same foibles and risks of human nature.

Cathy Bessant
Chief Operations and Technology Officer

As the power of AI continues to grow, political, business and social leaders will need to ensure that algorithms are endowed not just with “intelligence,” but also with the wisdom to understand and reflect the diversity of human culture and experience.

To do so, these leaders must work together to create firm and transparent governance processes. Amid consumer concerns about privacy and security, businesses can help ease fears about AI by being open about what data they are collecting, the reasons for collecting it, and how the data are being used.

Rather than resisting regulation and oversight, private business in particular have the opportunity to lead the way in creating strict ethical frameworks for the implementation and growth of responsible AI.

 

Ensuring equality, eliminating bias

Because algorithms are developed in the uncompromising realm of if/then logic, there is a misplaced optimism that they are immune to human faults, limitations, and biases. “That couldn't be further from the truth,” Bessant says. “As a product of humans, AI is subject to the same foibles and risks of human nature.”3
   
AI systems developed from a limited cultural, racial, or gender perspective cannot serve a diverse and dynamic society. A computer imbued with a narrow range of assumptions will not “learn” to be more inclusive—it will only repeat those assumptions at scale. Here, again, the solution comes down to humans, not machines. Committing to inclusivity and diversity offers the best opportunity to develop AI systems reflecting a variety of experiences and expectations.

Joining forces for a better future

The nature and scope of these challenges is such that no single actor can ensure that AI will be a force for good. To address these issues and help society as a whole find the necessary answers, in 2018 Bank of America became the founding donor of Harvard’s Kennedy School of Government to form the Council on the Responsible Use of Artificial Intelligence.
   
The Council is bringing together leaders from private business, education and government on a multidisciplinary approach aimed at capturing the growth potential of AI while taking a deliberate approach to minimizing the risks. “Our legal and judicial structures have no charted path for a lot of this,” Bessant notes. “We need to ask the right questions and see to it that deployment of AI doesn’t get ahead of the structures and the infrastructure needed to support it.”4
   
As we consider the future, the question is not whether we should continue with AI or retreat to a pre-AI world. If human history tells us anything, it’s that knowledge will surely advance, and so will the awesome capabilities of artificial intelligence. “At the end of the day,” says Bessant, “We all have a stake in this.”5

12/11/18


Open Location
Open How we're involved