Enterprises that want to accelerate their adoption of artificial intelligence should prioritize the most feasible projects with the greatest impact and take a “combinatory approach that includes analytics and the Internet of Things, experts told Microsoft’s FutureNow event.
A day-long conference hosted in Toronto, FutureNow keynotes included Mark Skilton, a professor at Warwick Business School and co-author of The Fourth Industrial Revolution: Responding The Impact Of Artificial Intelligence On Business. He said the race to to take advantage of technologies like natural language processing and machine learning needs to be balanced with a sense of what’s in the best interests of society, along with the expectation that many of the biggest breakthroughs are still ahead of us.
“We are not ethically obligated to build AI, but we are ethically obligated to build it responsibly,” he told the FutureNow crowd. “Most organizations are still in ‘discovery’ mode, and that’s okay. We’re all in this journey together. This is a learning curve for human beings as for machines.”
According to Brian Carpizo, senior expert, operations and advanced analytics at consulting firm McKinsey, the biggest use cases for AI at the moment call into enhancing operations, innovating new kinds of products and developing new business models. While the latter two categories tend to get a lot of the attention, he said industrial firms and health care organizations are already focusing on AI for operations enhancement.
“It’s not just a matter of (using AI for) preventative maintenance — it’s looking at the overall health of a machine, determining its useful lifespan and assessing how degradation of certain parts will affect the overall cost to maintain it,” he said.
A lot of AI isn’t far removed from advanced analytics, Carpizo added. The technology will help organizations predict things like demand for products and services, and also prescribe actions to change pricing of products and reduce customer churn. Besides keeping ROI top of mind, he suggested business leaders also look at where AI can help deal with human limitations such as productivity, cognitive bias, intellectual capacity and what he called ‘variability,’ where employees perform a task, such as assessing warranty claims, in an inconsistent manner.
“You have to think about how AI is going to change the way decisions happen,” he said. “When people start to say, ‘I’m 80 per cent sure this is going to happen’ how do you deal with that?”
Skilton recommended FutureNow attendees structure their AI projects by mapping what he called the “task environments” within their companies. These are the various operational areas of a business and the processes associated with them.
Next, it’s a matter of getting data ready, he said. This includes distinguishing what’s “learnable data,” which can be addressed by machines and some way versus “unknown data” that has not yet been properly defined or classified for AI systems. At that point developing the right solution can begin, he said.
Carpizo echoed Skilton’s sentiments, noting that besides weaving in technologies that can complement AI like analytics and the IoT, companies need to look at what’s truly possible versus what will have the biggest ROI.
“You could brainstorm 50 possible use cases,” he said. “But some of these things you just can’t do yet.”
FutureNow wrapped up Thursday.
Latest posts by Shane Schick (see all)
- FPX research shows 58% of B2B firms pursuing digital transformation want to improve buying processes - January 16, 2019
- Ethics guidelines from INFORMS focus on analytics and operations research - January 15, 2019
- CES 2019 experts identify enterprise XR opportunities based on virtual and augmented reality - January 11, 2019