Artificial Intelligence
Oct 04, 2017

Reflections from the 2017 AI Summit: Getting beyond the hype

After spending two days at the 2017 AI Summit in San Francisco, billed as the world’s number-one AI event for business, it’s hard not to get caught up in the AI hype. In fact, research firm Gartner places deep learning at the absolute peak of ‘inflated expectations’ in its latest Hype Cycle for Emerging Technologies.

During the summit, tech-industry speakers spoke about their experiences and expectations for AI. The line-up featured heavyweights Google, Microsoft, Facebook, Amazon, and IBM; venture capital firms, such as, Andreessen Horowitz; service providers and consulting firms, including Genpact, KPMG and PWC; and enterprises, such as Wells Fargo, American Airlines, Bristol-Myers Squibb, and Philips.

This post provides my reflections of the AI Summit, and recommendations for companies exploring AI.

AI as the new electricity

Andrew Ng, co-founder of Coursera and former chief scientist of Baidu: “Just as electricity transformed industry after industry 100 years ago, I think AI will do the same.”

Sometimes technological change is so compelling that it drives significant economic, business, and social change. AI is one such technology. That said, while there is the CNN view of AI (i.e., fears of AI immediately replacing people’s jobs and creating significant longer-term risks for all of civilization), there is also the more pragmatic view. I’d urge enterprise leaders to take a pragmatic view, which can be hard to see through the hype.

All participating tech titans stated how AI is strategically important for their companies. Google spoke about “pushing the boundaries of machine learning” and described itself as “an AI company.” Microsoft’s mission is to improve the productivity of individuals and companies, and it sees AI as “human ingenuity amplified by intelligent technology.” In other words, AI is a key ingredient to realize its vision. IBM, not surprisingly, talked up Watson and how it believes it can transform many industries.

Frank Chen, partner at Andreessen Horowitz, was extremely bullish on AI. He provided many examples of how the firm’s portfolio companies use AI to disrupt entrenched competitors. Textio, for example, has AI-tuned job descriptions to enable better hiring. And, Everlaw is transforming legal e-discovery work with AI technologies. He also described many AI-based capabilities that we will see in the near future, such as autonomous vehicles driving beer from breweries to retail outlets, and drones delivering it to our homes.

Several enterprises described their upbeat views of AI. Johnson & Johnson said AI will fundamentally change the future of healthcare. American Airlines described how it is planning to use AI to transform its customer experience within airports. And Netflix talked about how it uses AI to select the movie images presented to over 100 million consumers to increase the probability that they will click and watch.

AI hype might lead us to believe that all companies are currently working to deploy some type of AI-based solutions. While recent research by Genpact and Fortune Knowledge Group shows that 82% of organizations plan to implement AI-related technologies in the next three years, only 25% are generating significant positive business outcomes from AI today. Not surprisingly, AI Summit organizers selected their enterprise speakers who are in the minority.

Humans key to putting AI to work in enterprise

Steve Jobs: “Technology is nothing. What’s important is that you have faith in people.”

Although AI Summit stories may lead us to believe that technology, data, and algorithms can do it all, this is an AI myth. People matter. Technology advances relentlessly, but to apply it successfully within your company’s environment, the industry and functional knowledge people have is critical to understanding the context and goals for your project.

This knowledge provides the context that creates value from data by turning it into useful information. People understand the right questions to ask. We understand the purpose of a project. We understand the customer experience we want to create.

One AI Summit presenter from CrowdFlower put it this way: “AI = training data + machine learning + HITL, where HITL = humans in the loop.” In other words, successful AI requires people’s domain expertise. So yes, people do matter and the companies that come out on top will be the ones that combine human knowledge with AI. 

To make this combination effective, companies must also reskill their people. Surprisingly, out of the 82% of organizations planning to apply AI in our study, only 38% say they currently provide employees with reskilling and training opportunities in the face of technology disruption.

In short, to succeed with artificial intelligence, companies should consider a number of key elements. If you don’t have the capabilities you need in-house, use knowledgeable service providers. Since AI has the potential to fundamentally change your business model, it is a board-level strategy topic, not merely a technology play. Buy-in from the executive team is key. Get the governance right from the start. And, last but not least, begin with the end in mind.

About the author

Dan Glessner

Dan Glessner

Vice President, Digital

Follow Dan Glessner on LinkedIn