Even considering all the buzz created in the last years around the implementation of artificial intelligence by business corporations, globally speaking, it is already possible to perceive some emerging realities seen in the real economy. A recent research published by MIT Sloan Management Review in partnership with BCG Henderson Institute points to some of these directions and serves to help guidance for those interested or involved in this industry.

The research was conducted with executives from 29 different industries and located in 126 countries. Since more than two thirds of respondents were from outside the U.S., it seems to be not so biased towards North American management practices and reflects different approach from Chinese companies, just as an example. Most important, a simple analysis of the report demonstrates that it is almost mandatory to divide respondents’ organizations in subgroups, due to the maturity level of understanding of AI concepts and, of course, the standards of implementation of AI applications.

The Pioneers, corporations that both understand and adopted AI solutions, are only 18% of the respondents.These companies are actually leading the track towards implementing AI solutions not only to their products or services offerings but also to their internal processes. Investigators, the second largest group with 33% of respondents, are those corporations investigating various possibilities, but with low adoption of AI applications. The Experimenters, with 16% of the respondents, are deploying AI solutions without deep understanding, learning by doing in a expected to be virtuous cycle. The largest group are the Passives, with 34% of respondents, those with no adoption or not much understanding of AI implications in their business.

What can be immediately inferred by these subgroups descriptions is that more than 80% of the respondents’ organizations have low or no adoption of AI applications, and any analysis of success cases would need to be focused on the Pioneers. Putting it in context, each industry particulars and organizational culture play a huge role on this game, but what all of them have in common – fierce competition – seems to be the sparks igniting the flame of innovation. Very possibly, these cases will serve as benchmark for the other groups in the near future – historically, something very common in the adoption of new technologies.

A very large group of Pioneers respondents’ organizations, 80%, see data as a very important corporate asset and, especially important, on the agenda of senior management. So, it is not a coincidence that Pioneers respondents confirmed the existence of an AI strategy, with a solid 90% answers clearly stating the importance of the use of AI in their business. And 69% of them declare the growth of their understanding about AI applications use in comparison to last year, suggesting that early initiatives have been encouraging. The reality is that corporations can learn about AI in various different ways: Pioneers informed they are investing more in the past year than in prior years in AI talents and AI technologies, like the quality of data required to train AI algorithms and the quality of processes required to train the algorithms.

Relatively to their peers from other parts of the World, Chinese corporations are aggressively investing in AI applications. And here comes the first competitive advantage: since AI applications learn by the use of training data – the more the better – Chinese Pioneers centralize the housing and governing of huge amount of data in a very competent manner.  As a matter of fact, 78% maintain their corporate data in centralized data lakes, in comparison to 37% of Europeans and 43% of U.S. Pioneers. And 83% of Chinese respondents manage corporate data centrally, while only 39% of European and 40% of U.S. Pioneers counterparts do so. Also interesting, it seems Pioneers from everywhere are eager to scale AI throughout their enterprises, but whereas Chinese corporations are more cost-savings oriented, the rest of the World seems to have a major focus on revenue generating applications.

In a typical two-speed process, Pioneers seem to try to meet their current business needs, while trying to focus on core process models. In other words, this group is adapting their strategy to build processes and platforms for AI at scale which will not be a simple task. Organizational culture is a complex issue by itself and driving AI understanding and action into business will require efforts from individuals, corporations and societies. In this particular case, the research suggests managers recognize both the opportunities and risks of using AI – to different degrees. Curiously enough, 39% of all respondents consider AI a mixed blessing, reflecting a recognition that the same AI advantage a corporation is looking to launch becomes a major risk if a competitor moves faster and deploys first.

It is proven already that AI is not a just a lab experiment, it is actually adding value to real corporations. Even though the number of Pioneers are not increasing in comparison to the other groups, it is also a fact that they are investing more in AI, therefore widening the gap with the others. It is also true that all groups face different challenges like how to take AI to industrial scale in at least three dimensions: leveraging internal processes, reducing production and delivery costs and finally, generating revenues.