Quantopian is not your usual hedge fund. The Boston-based company has built a platform bringing together freelance quantitative analysts to create trading algorithms. Everything is completely automated, with complicated mathematics replacing old-fashioned financial nous. The platform’s value proposition lies in its own users and their work, rather than a specific product.
If Quantopian is currently an outlier in the world of finance, our tech-obsessed culture will make it the norm, argue Andrew McAfee and Erik Brynjolfsson in their latest best-seller Machine, Platform, Crowd. As with many other futurologists, the two MIT academics anticipate an avalanche of disruption sparked by artificial intelligence (AI), as soon as “in the coming decade’’. Robots and algorithms are expected to upend every sector, from manufacturing and finance to areas hitherto immune to automation, such as law and education.
The authors are no strangers to doom-and-gloom predictions. Having made their name with Race Against the Machine, a pessimistic account of the impact of automation on blue-collar jobs, they now take the argument one step further. The key idea is that as machines get smarter, our ability to compete with them in crucial tasks such as quantitative analysis will dramatically diminish.
The implications for business leadership are immense. Forget old-fashioned Mad Men-style executives who based strategy on intuition, commercial nous or experience. Rockefellers and Fords will be no more. Algorithms will be calling the shots, pointing to the right direction for the organisation. Humans, argue McAfee and Brynjolfsson, should focus on less prosaic tasks, such as figuring out what problems machines should be working on; AI might be a great tool to provide solutions, but they will never be as good as humans at posing the right questions, since they don’t live in our world.
If this sounds like a slightly dehumanising process, striping executives of their hard-won powers and status, McAfee and Brynjolfsson beg to differ. There will always be a role for them as storytellers: selling ideas to the crowd, which itself will be playing an increasingly important role on digital platforms, as testified by the success of Quantopian.
Managers may have to counterbalance the cold thinking of machines with a soft, compassionate approach, says Dr Milo Jones, Visiting Professor at IE Business School and expert on the impact of technology on business: “The shift from macho strategy towards smooth processes, storytelling and emphasis on fairness, or at least appearance of fairness, will be important. AI will also be amoral; we will try to build ethical decisions into it, but business leaders will need to act as moral guardians of decisions. All that suits a more feminine leadership style, at least in terms of 20th century clichés.”
Even those invaluable social skills will be enhanced by technology, says Dr George Zarkadakis, Digital Lead at the professional services firm Willis Towers Watson and expert on the impact of AI on organisations: “Mentoring people will remain a human endeavour, as well as facilitating collaboration in teams; but for both these aspects of leadership embracing AI will definitely be a prerequisite, for one will need AI for augmentation.”
Centaurs can be choosers
Not everybody subscribes to the idea that the robots have already won the boardroom battle. Many experts propose a more balanced synergy of humans and machines; Jones calls this the “centaur model”, inspired from the half-man, half-horse mythical creature.
Jones has borrowed the term from the epic “freestyle chess” competition the Russian champion Gary Kasparov organised after losing from Deep Blue, an IBM chess-playing computer. Deeply annoyed by his defeat by a machine “as intelligent as an alarm clock”, as he put it, Kasparov organised a chess competition where teams of chess geniuses partnering with high-tech computers competed against each other; Kasparov called them “centaurs”. Surprisingly, the winning team comprised of two medium-skilled chess players with three ordinary Dell computers. After studying their methods, Kasparov figured out that even conventional human and artificial intelligence can produce astounding results, as long as it follows sound processes to figure out which decisions should be made by humans and which ones should be left to machines.
This is an apt metaphor for the future of leadership, as process will gradually trump grand strategy, says Jones: “It’s about figuring out what humans are good at, what we know about competition, what factors the software doesn’t account for, but also when to step back and trust the software.” In a way, this era is already here, says Jones who often works with commercial powerhouses such as Walmart: “We now live in a world where teams of machines and humans compete against each other, even in business. So everybody is a ‘centaur’ now.”
Machines have other inherent limitations that might be difficult to tackle. As Jaron Lanier, a virtual reality pioneer, argues in his latest book Dawn of the New Everything, robots will be as good as the data we feed them with. That will always limit the range of tasks they will be able to perform. And if there is one skill that cannot be artificially replicated, it’s the ability to connect the dots, thinks Jones: “What computers can’t do is synthesis: the ability to bring disparate ideas together. If I were in the sugar industry, I should be thinking about the problems facing the tobacco industry because people might now raise health concerns about my business. It’s that kind of analogical thinking that computers are nowhere close to.”
Organisational structures may change too. McAfee and Brynjolfsson suggest that the knowledge economy favours egalitarianism and transparency, stopping short of embracing the flat hierarchy model that reigns supreme in Silicon Valley. No longer will unchallenged authority, what they call HiPPO (the Highest Paid Person’s Opinion), dominate decision-making. That era may be gone anyway, says Zarkadakis: “Most organisations nowadays follow a shared leadership model, which makes more sense in dealing with complex problems in a complex world. AI will help to reduce much of this complexity by taking care of many analytical tasks that business leaders still find themselves doing.”
What’s more, AI-driven companies may even kill grand strategy according to McAfee and Brynjolfsson. AI will enable companies to shift from away from long-term forecasting and big bets to short-term experimentation. However, this might be difficult in areas such as finance, where executives are often accused of ‘short-termism’: prioritising quarterly results over long-term goals. Many businesses may also have a hard time replacing traditional decision-making processes with tech-driven ones, says Dr Will Venters, Assistant Professor of Information Systems at LSE: “Organisations underestimate the complexity of their business processes and therefore the difficulty of imbedding any new technology in them, particularly technology taking over intelligence tasks.” Business will always need a coherent mission, says Zarkadakis: “The real problem with fail-fast experimenting, agile methods is that they often produce outcomes that do not align with strategy. Business needs a strategy to get alignment across teams and for benchmarking, and I do not see a time when strategy will be replaced by R&D going wild…”
From Bill Gates to Steve Jobs, all great managers of our times shared a common trait: an in-depth understanding of technology. How much important will this be in the era of omniscient robots? A lot, say McAfee and Brynjolfsson, recommending ‘geeky leadership’, defined as “technically proficient leadership”, as the way of the future.
Geeky leaders are connoisseurs of technology, in many cases developers themselves, a trait that bestows them with respect within tech-driven organisations. As great leaders of the past, they are visionaries who know where they want to go. But they also have their feet on the ground and articulate their dreams in an unassuming way to take enough people on board. They often have strong opinions too: in the AI boardroom there is no place for creative ambiguity and flowery language.
Some dismiss the idea that tech savvy will be a prerequisite for successful leadership. “It’s far more important to be able to bring free-floating, humanistic knowledge to management rather than technical skills,” says Jones. To highlight his point, he uses an example from the previous wave of technological innovation: “Cars used to be so complicated that chauffeurs were needed to drive them, but now everybody drives their own car. Studying too much about technology is the equivalent of training to be a chauffeur in a world where computers will be increasingly driving themselves.” A minimum of tech savvy could be helpful though: “Vendors will always tell you to buy the most expansive and powerful software, but you will need to have an intuitive understanding of what is usable and what is not, what is the best way to display data, which software has the best human-computer interface. You have to know enough so that people can’t fool you.”
However, even those business leaders who combine a creative mind with a technical background may have to cede some of their authority to machines. Those who fail to adapt will face oblivion. As McAfee and Brynjolfsson put it: “Over the next decade, AI won’t replace managers, but managers who use AI will replace those who don’t.’’