Erik Brynjolfsson
Stanford HAI
In a 2018 Financial Times article, The Long Wait for a Productivity Resurgence, it was noted that “No economist has done more to promote the revolutionary implications of information technology than MIT’s Erik Brynjolfsson”. Brynjolfsson was then professor at MIT’s Sloan School of Management, a role he held for 20 years, while also serving as director of the MIT Initiative on the Digital Economy (IDE). During his two decades at MIT, Brynjolfsson distinguished himself as one of the world’s leading researchers on productivity and work, and on the economics of information, AI, and the digital economy. An economist by training, Brynjolfsson has focused on how AI drives productivity. When it comes to the business productivity implications of AI, who better to provide a perspective on the current state of AI and its potential to transform work.
I first met Erik Brynjolfsson during his tenure at MIT. At the time, I attended a number of IDE events, and Brynjolfsson joined me for one of a series of Fortune 1000 CIO roundtable tables that I organized and hosted during these years. When I published my book, Fail Fast, Learn Faster: Lessons in Data Driven Leadership in an Age of Disruption, Big Data, and AI, in 2021, I referenced Brynjolfsson’s pioneering work at the intersection of digital, data, and AI, which was years ahead of its time. In works such as 2016’s The Second Machine Age, co-authored with Andrew McAfee, and in a 2017 MIT Sloan Management Review article, “How to Thrive – and Survive – in a World of AI Disruption”, Brynjolfsson anticipated the impact that AI would have on work and business productivity.
In June 2020, during the COVID outbreak, Brynjolfsson was presented with the opportunity to relocate across the country to become the Jerry Yang and Akiko Yamazaki Professor and Senior Fellow with Stanford University in Palo Alto, and to launch the Digital Economy Lab at the Stanford Institute for Human-Centered AI (HAI). The mission of the Digital Economy Lab is to improve our understanding of the economic implications of the rapid improvements in AI and other digital technologies. Led by Brynjolfsson, it includes faculty from multiple departments across Stanford. Research focuses on developing new economic models and empirical research, including measuring, forecasting and guiding the human and societal impact of AI, and particularly in ways that create widely shared prosperity.
I had the opportunity to catch up with Brynjolfsson in recent weeks. Brynjolfsson’s research examines the effects of not only AI but also other information technologies on business strategy, productivity and performance, and digital commerce. He believes that the impact of a changing economy will be felt by everyone, crossing boundaries of wealth, status, geography, and culture. Central to the mission of the Lab is a belief in the power of human work, and a belief that augmenting human capabilities with machine capabilities through AI will lead to beneficial gains for everyone. As Brynjolfsson noted as far back in 2017 “AI won’t replace managers, but managers who use AI will replace those who don’t.”
I asked Brynjolfsson how it felt to relocate to the West Coast and in close proximity to Silicon Valley. Brynjolfsson expressed deep love for MIT but added that it was invigorating to be near and a part of such a strong innovation community, both at Stanford and in Silicon Valley. I noted that I had recently had the opportunity to hear HAI director Fei-Fei Li speak at the Wall Street Journal Tech Live event in Laguna Beach and had written about that experience, and been inspired to read her memoir about her life in AI. HAI has assembled a strong leadership team and distinguished fellows, including the recent announcement that noted MIT researcher Sandy Pentland would be joining the HAI team as part of Brynjolfsson’s Digital Economy Lab.
Brynjolfsson recently participated in an extensive interview with the Financial Times, which appeared under the headline, “Erik Brynjolfsson: ‘This Could be The Best Decade in History – or the Worst”. I asked Brynjolfsson about his perspective on AI in the context of history, and where he would place AI among the pantheon of transformative technologies. “I think it’s most comparable to the Industrial Revolution” commented Brynjolfsson, noting however that, “while the Industrial Revolution played out over more than a century, AI will play out much faster, in a matter of a few decades”. AI is coming along “faster than anyone really expected” added Brynjolfsson, citing AI in healthcare as an area of great potential for human good.
I noted that in his Financial Times interview, Brynjolfsson had commented that people often use AGI – artificial general intelligence — as a synonym for human-level intelligence. However, “I think it will become apparent that humans have a narrow kind of intelligence and that truly general intelligence will have a much broader set of capabilities”. This was not something I commonly hear, so I asked Brynjolfsson to elaborate. This led to a discussion of augmentation, and how AI can and should be used to augment human skills and intelligence, which is at the heart of driving improved productivity. Brynjolfsson explained that technologists need to look at how AI can be used not just to “imitate or repeat human tasks”, but how AI can be used to “augment and complement” employees.
As an extension of his commitment to improving productivity, Brynjolfsson is embarking on a new business venture focused on “task-based analysis”, which refers to the concept of looking at tens of thousands of tasks and ranking them based on which tasks AI can help the most with. Brynjolfsson has launched a venture, Workhelix, which combines research and data to quantitatively assess a company’s GenAI opportunity. Workhelix co-founder and CEO James Milin explains, “The approach that we’ve pioneered starts with a granular question: can GenAI double the productivity of a task without reducing quality?” He continues, “Using our algorithms and data, we answer this question for every task in an organization, then work upward to understand how jobs, and ultimately each company, can benefit from GenAI”.
Brynjolfsson is a powerful advocate of the productivity potential that can be achieved through AI, envisioning double-digit gains in economic productivity, and commenting that he foresees the potential for “massive economic disruption” leading to the creation of new occupations and new companies in the coming years. Brynjolfsson believes this can lead to greater social benefits, manifested in the form of higher wages and greater prosperity. He expands upon this thesis in an article, The Turing Trap: The Promise and Peril of Human-Like Artificial Intelligence, noting that, “When AI is focused on augmenting humans rather than mimicking them, than humans retain the power to insist on a share of the value you created”. He adds, “Augmentation creates new capabilities and new products and services”.
As with any technology and period of historic transformation, responsibility ultimately resides with us as human beings. Brynjolfsson has argued that because we have more powerful tools than ever before “this could be the best decade in history – or the worst”, noting that “misinformation, viruses or weapons, cyber-attacks, and phishing attacks” remain threats. He points to risks posed by information overload, fake news, and other nefarious uses. Brynjolfsson concludes that the future of AI will largely depend on the steps we take today as humans to place a premium on intelligent governance, safeguards, and guardrails. Brynjolfsson warns us simply, “We should not fly blind.”
