Most of us don’t tend to have very good data about the future, so predictions are hard. One of the few ways to use data for predictions is to extrapolate from surveys about what early adopters are doing. As the science fiction author William Gibson famously noted, “The future is already here—it’s just not very evenly distributed.” If we look at what bellwether companies are doing—like the belled sheep that lead their flocks to and from pastures—we can get a sense of what will happen at other firms over the next few years.
In 2017 I worked with Deloitte on a survey of early adopters of AI or cognitive technologies. It wasn’t easy to find them; we started with 1507 U.S. managers, and eventually found 250 whose organizations were doing a lot with AI and who were familiar with their firms’ activities in the area. They also had to be senior executives from fairly large organizations, which also reduced the numbers. We called it the “cognitive aware” survey, although the respondents could also be called a “cognitive active” group.
These companies suggest several likely directions for enterprise AI over the next several years. They include:
Substantially more investment and application—Large majorities (in the 90% range) felt that AI was important or very important to both product and service offerings and internal business strategies. 91% agreed that AI is a strategic priority overall for their firms. About 60% said that AI was both the technology driving change most and receiving the most investment in their businesses. Based on that level of importance, it’s likely that other companies will follow along for both internally- and externally-oriented applications.
Business transformation from AI is coming…soon—I tend to be a bit conservative about how long business transformation from any technology will take, but these respondents weren’t. 76% said that AI would transform their companies within at least 3 years; only 3% predicted it would take 10 years or never happen. 16% said it’s already happening. And these executives were confident about their companies’ abilities to change: they predicted faster change for their own firms than for their industries overall. That’s probably fair, since they were early adopters in the first place.
We’re in for some interesting new products and services—I expected that the primary objective of AI in these companies would be internal process optimization, but “enhance the features, functions, and/or performance of our products and services” was the most popular objective, with 51% saying it was their first, second or third leading objective. “Create new products” and “pursue new markets” weren’t far behind, with 36% and 25%, respectively, putting them in the top three objectives. These efforts seem likely to result in some highly intelligent new products and services, and I can’t wait to see them.
It’s unlikely we’ll see a lot of job loss from AI—On that question of objectives that I just discussed, “Reduce headcount through automation” was the lowest-ranked of all objectives chosen, with only 22% putting it in their top three objectives. And most respondents in the survey also said they do not predict substantial job losses. Within the next three years, 69 percent of enterprises anticipated minimal to no job loss and even some job gains from AI/cognitive. In ten years, 22% did predict substantial job loss, but a larger number, 28% predicted many new jobs from AI in that timeframe. Neither the respondents nor we know the net job outcome, but these bellwether firms don’t seem too concerned about massive unemployment.
Get ready for job change and retraining—Jobs may not disappear, but they will certainly change according to these “cognitive aware” executives. 56% saw substantial or moderate change in job roles and skills now, and 62% anticipated it within three years. 64% said their organizations would need substantial training or retraining to learn how to work effectively alongside smart machines. Somewhat surprisingly, 63% said they already had some form of retraining underway involving AI. I haven’t seen that many programs in the companies I work with, but maybe I am hanging out in the wrong ones.
In short, these bellwether respondents anticipate a lot of AI, a lot of business transformation, and a lot of jobs involving a lot of change. Of course, they could be wrong; the data from any such survey is subjective. But it’s also worth pointing out that the sub-group of the respondents with the greatest level of experience with AI/cognitive tended to be the most positive about the current and future impacts of the technology. I’ve been working with information technology in large organizations for 35 years, and I don’t recall so much enthusiasm for any technology in that time.
