The artificial intelligence race grabs headlines and the narrative seldom varies: On the world stage, it’s U.S. innovation versus China’s speed. In business, senior leadership at all sorts of organizations are romanced by the prospect of winning the AI race. Some believe it’s just a matter of greater investment and a policy change or two.
This begs the question: Is AI a race to be won?
A race is a contest to determine who is the fastest at covering a set course, one with a defined start and finish line. Obstacles are identical for competitors because, of course, a race must be fair with the same rules for all.
Get there first and you’ve won. The winner accepts the trophy and holds the title—and a Wikipedia entry—in perpetuity. Take the case of the first automobile race held in the United States in 1895. Six motorized vehicles participated, four cars and two motorcycles, and they raced from Chicago to Evanston and back. Frank Duryea’s motorized wagon won the race, and a Benz car came in second. Duryea went on to build 13 cars, a feat in those days—but today, all over the world, we still drive Benz cars. Some journeys in life extend beyond near-term milestones. AI in organizations is in a similar vein, I think.
Let’s revisit the idea of a race in the context of AI in an organization. AI is a capability that enables us to do our most important tasks better than before. It can help make recommendations to our sales force on the next-best action for a customer, identify defects on the manufacturing floor, better optimize inventory in the supply chain and interpret images and text as humans do. And, in many cases, we increase our capabilities by using many open-source algorithms that are available to all.
AI is not a singular technology that we individually own because we got there first. If we find clever, novel ways to forecast, that does not mean that others cannot find smarter methods as well. In some cases, AI may be a product that we get to design first, and perhaps even patent, but getting there first doesn’t stop others from getting there, too. As digital proliferation increases, the state of the possible is well understood by all. It’s no secret that online ordering/curbside pickup is fueled by AI. The actual algorithms (how data is stored and utilized) may be a trade secret, but the use case is not.
There is no endpoint, no finish line. As an enabler, AI needs to keep getting better. The processes keep evolving. AI’s goal is to mimic human intelligence and help us improve. As customers and markets evolve, organizations need to be creative and keep refining to stay competitive. Some argue that with specific AI applications, such as autonomous cars or, more broadly, autonomous systems, there is economic incentive to finish first or near the top. That may be—and indeed, there is some glory and front-runner status to be the first to beat the human Go champion or other equally daunting challenges—but that does not preclude others from working to improve their capabilities on everyday tasks that can be benefited from AI. So, identify your most important problems that can benefit from AI and find ways to get better. Your competition is the present way of doing things.