Delivering boring AI with Decision Management
Blog: Decision Management Solutions - Decision Modeling
A great article appeared in Information Age recently based on an interview with Tom Davenport If you want to see the benefits of AI, forget moonshots and think boring. In it, Tom argues that “if enterprises ever want to see the benefits of AI, they must embrace the mundane”.
This is particularly true for companies that aren’t tech startups. As Tom says, “moonshots are possible if you’re a tech giant and you have billions of dollars to spend on experimenting” but what if you can’t “pivot”? What if you have to keep doing what you have always done – manufacturing things, selling insurance, providing banking services? Big, boring companies need to take a different approach. They need to find practical, operational decisions they make today and figure out how to apply AI to make those decisions more profitably.
This is the key for success we see in clients with AI. Those trying to bite off huge problems like “should we pay this claim” with a single AI algorithm are managing to spend a lot of money but seeing very limited results. Those that break their business problems down into smaller pieces find lots of practical use cases for AI such as “how likely is this person to have an undisclosed medical condition” or “does this doctor’s note describe the condition the claimant says they have”.
To succeed with AI, understand how you decide today. Build a model of your decision-making approach to get everyone on the same page. Then ask how AI can help. “If only we knew XXX, we would decide differently. Can AI help us know that?” Integrate these smaller, practical use cases for AI with business rules and other analytics to deliver real business value. It costs less and works better.
Remember, even Jeff Bezos admitted the bulk of their AI investments are “quietly but meaningfully improving core operations.” Decision Management is how you can do the same.
Contact us if you would like more information on how we can help you do this.