Organizations considering AI investments should “think big, start small,” according to Tom Davenport, the President’s Distinguished Professor of Information Technology and Management at Babson College. During his address at this year’s Digital Workforce Summit, Davenport urged companies to master basic use cases rather than chasing moonshots.
“Think big, start small” is how organizations considering Artificial Intelligence (AI) investments should proceed with any project, according to said Tom Davenport, the President’s Distinguished Professor of Information Technology and Management at Babson College. During his address at this year’s Digital Workforce Summit, Davenport urged companies to attack and master basic use cases before attempting moonshots.
Davenport, who is also the co-founder of the International Institute for Analytics and a Senior Advisor to Deloitte Analytics, has written or edited twenty books and more than 250 print or digital articles on AI and other advanced technologies.
When beginning research on AI strategy, Davenport urged companies to take “your strategy, your business model, your business processes and think about how AI can transform it. But develop a series of smaller projects, that's what AI does really well anyway. It does small things really well, and if you do a number of those small things, you can transform customer service, for example.”
As businesses find success automating simple tasks, they should scale up and “skill out.” In other words, increase how much AI automates, but also increase the breadth of tasks AI is capable of automating.
Contemporary AI Use Cases
Davenport said he works with a lot of companies focused on product-related objectives, such as creating new products and services with AI. He also said companies have used AI to add capabilities to existing products. Additionally, companies are using Machine Learning and analytics to optimize internal operations and make better decisions.
He referenced Amazon’s use of AI as an interesting example of how to apply AI technology to existing business models. With Amazon Go stores, drone delivery and other AI-based applications, Amazon has taken the area where it already excelled — selling and delivering products — and made itself even better.
“[Amazon CEO Jeff] Bezos said, ‘The bulk of what we do is quietly but meaningfully improving core operations,’ and that says to me, bit by bit by bit, we're making this company more effective,” Davenport said about Amazon’s AI evolution. “The low-hanging fruit, even at a place like Amazon, is paying off quite well.”
AI and a Hybrid Workforce
AI will not enable companies to simply replace employees with automation, Davenport said. In fact, he referenced the Amazon Go stores as a prime example of how humans and AI complement one other well for tasks related to improving customer experiences.
“There still are, by the way, humans in those [Amazon Go] stores, [including] someone to show you how to use the [Amazon Go] app,” he said. “They weren't quite sure in the Seattle store that I was over 21, so they had to have me show [a human] my license when I wanted to buy a bottle of wine. And then somebody broke their wine, not me, and cleaning it up took another human being, so that's three people [working] in a convenience store.”
No matter how simple a company’s AI use case will be, it will still require overcoming a bit a trial and error. Davenport cited implementation, scaling, and re-skilling workers as challenges for companies attempting to manage new AI-augmented processes.
“We need to make people ready for this, and not enough organizations are doing that yet,” Davenport said.