Cultural Characteristics for Adoption of Machine Learning and Artificial Intelligence in Large Organizations

Building an adequate organizational culture helps with artificial intelligence and machine learning adoption. Limited research on how to transform cultures appropriately exists. Discover how you can contribute to building a more suitable workplace for people at your firm.

SwissCognitive Guest Blogger: Jean Voigt, Previous Head Portfolio & Capital Analytics at Credit Suisse

Did you ever end up at a dead-end right in front of a wall and wondered how did you ever get there?

These days organizational culture and AI are two fields with lots of attention. To start with, AI and ML get confused with each other, despite limitations and learning being a pre-condition for intelligence. Even though it may seem, at times, there is more talk than walk. Lost in a great conversation, a casual walker may quickly end up at entirely the wrong place.

Several firms are at risk of going down the wrong path when organizing their AI or ML initiatives. This article attempts to go beyond the mythical experience account of what works and highlights the need for more data to establish supportive organizational factors. It is a call to action to make the debate a little more objective.

After years of experience building AI-enabled teams and organizations, I can share plenty of observations. However, these are all my very own perspective and lack validation by data. It may well have been that my remarks were shaped more by circumstances than anything else. It is long overdue to start collecting data on the matter.

A few weeks ago, the SwissCognitive re-initiated a debate that has fascinated academics for some time and centered on the human-AI nexus. Observations in practice confirmed that trust and empowerment reduce the sensation of loss of control and seemingly aid AI adoption. However, other aspects, such as failure-positive environments, seem necessary as well. Especially for AI and data teams because that is at the root of agile and data-driven transformation. But what else is helpful?

For large organizations being failure-positive is a challenge, not only because of the quarterly earnings call. For the sake of harmony, large organizations can afford to hang on to poor practices for longer than smaller firms.

It appears that even technology companies building AI products suffer from cultural challenges when they mature. How should less technical and mature organizations design and facilitate their culture to drive healthy AI and ML adoption, especially without the vast resources of large technology companies?

AI in general, and machine learning specifically, is technical. The organizational setup to facilitate the adoption seems not very well understood, despite many “success guides” in every form and shape. Making matters worse, there is a constant underlying fear of AI. Prescriptive guidance of the “correct” steps to take or the “right” set of roles falls short of recognizing the specific nature of individual firms. Abstraction helps. Rather than looking to particular elements, principles in the form of organizational culture provide an excellent basis to start the journey on the product you build for your people.

In an attempt to combine exploratory hypotheses on AI and ML adoption and beneficial cultural characteristics, I like to cordially invite you to participate in a survey on that matter.

Despite past research inside selected organizations such as Google or Microsoft, few data points exist across companies outside the technology industry. Many focus on structural aspects rather than cultural. Even less research covers multiple companies, despite some work on the country level. Moreover, what works at Google, may not work at your family-owned 150 people company.

Based on research regarding corporate culture characteristics, the survey asks for respondents’ opinions regarding supportive cultural factors of AI and ML adoption. In addition, several facilitating activities hypothesized previously are available to respondents’ ratings. Results are collected anonymously and will be published at SwissCognitive after reaching a sufficiently large sample size.

I am hoping that this survey may contribute important insights into cultural and AI organizational design. Not just talk, but some facts to guide the way for organizations plotting their path through this unwieldy territory. By sharing and responding to the survey, you contribute to fewer organizations making expensive mistakes when building AI and ML teams.

Click for the Survey

About the author:

Jean has been building and leading data, artificial intelligence, and machine learning teams for more than ten years. As a CFA Charterholder with a Ph.D. in Computer Science from the University of Zurich, he has built expertise in driving the organizational design and strategic product management aspect of machine intelligence. He is the author of the Machine Learning Manifesto and Unmanage a book for executives and machine learning engineering leaders.