Potential Uses of AI to Develop More Inclusive Leaders

Here are ideas for machine learning tools that would help leaders recognize and address their unconscious biases and exclusionary communication habits.

It’s hard to find an executive who is not prioritizing developing a more diverse workforce and inclusive work culture. This is understandable given the overwhelming research about the effectiveness of well-structured diverse teams. However, corporations’ spending on diversity initiatives has utterly failed to move the needle. This is in part because their training programs fail to help leaders recognize and change destructive patterns in their personal communication and decision making. These patterns are expressions of unconscious bias that we all hold, and we require frequent feedback in order to address them.

As the NeuroLeadership Institute notes[i], “Whether you’re trying to become more innovative, agile, or inclusive, changing culture is really a matter of changing shared everyday habits.” As Christine Comaford wrote in Forbes[ii], “To truly promote diversity and inclusion, it is absolutely critical to train your team in effective communication skills.”

Fortunately, the Artificial Intelligence Revolution is providing us with new tools that can be used to help leaders identify and address destructive communication habits, assisting them with being more inclusive. Computers are capable of using machine learning to rapidly analyze our workplace communications to provide real-time feedback and post-meeting or post-presentation scorecards to executives and team leaders. The tools already exist — computers can analyze audio recordings of meetings captured by tools like the Amazon Alexa, they can analyze videos taken by videoconferencing cameras in boardrooms, and they can analyze our emails with simple text-based email scanning — they just need to be re-purposed.

Here are some ideas to get computer scientists’ creative juices flowing:

For the boardroom:

· A tool that buzzes your phone in real-time when you interrupt the speaker in a meeting. Taken to the next level, what if the tool could be programmed to selectively alert a male executive who is currently interrupting a female colleague?

· A tool that provides executives immediate feedback during the break in a brainstorming meeting about how frequently they used “yes, and” versus “no, but” style language in response to proposed ideas. This tool would be especially useful if the feedback was coded by whether the idea generator was of a particular race or gender.

· A tool that analyzes who the executive does and doesn’t make eye contact with during presentations.

· A tool that helps leaders recognize the types of speakers to whom they give time to express or formulate ideas.

For an executive’s office or for the company canteen:

· A tool that would recognize when an employee refers to a female colleague as a “girl” or a “chick” and sends a feedback email to that employee (the data shouldn’t be stored or shared with management — I’m not advocating for a Big Brother state, but we can use smart recording tools to help people become aware of communication patterns without having to bring performance management into the picture).

· A tool that helps executives learn who they raise their voices at or speak to with an aggressive tone, versus those with whom they are more calm and respectful with. Similarly, imagine a tool that helps leaders learn the types of conversation content or comments that trigger their overly-emotional or harsh reactions.

For scanning executives’ emails to provide anonymous feedback reports:

· A tool that gives an executive an analysis report of their emails to various subgroups in their departments. Is an executive expressing gratitude more frequently to people of certain races? Are they using more authoritative or authoritarian communication styles with employees of certain genders?

In addition to providing an executive with feedback about ways in which they are making their workplaces more or less inclusive through their communication, the same tools can be used to help coach rising leaders from underrepresented backgrounds to gain more authority and respect within the work place. For instance, AI tools can help women leaders recognize if they are too frequently apologizing for mistakes they didn’t make, or if they are overusing verbal ticks or fillers that hurt their communication quality. Similarly, machine learning can analyze verbal and nonverbal communication patterns to help rising leaders recognize how they are subconsciously expressing discomfort or lack of confidence in presentations or feedback conversations.

There is no shortage of ways that we can use artificial intelligence to create more inclusive leadership and workplaces. Some of the ideas listed above will require significant amounts of work to bring to reality, while others only require some easy adaptation of existing tools. I hope socially conscious entrepreneurs and computer scientists heed this call to action to use their talents to revolutionize diversity and inclusion training programs.

[i] https://neuroleadership.com/solutions-for-organizations/culture-leadership

[ii] https://www.forbes.com/sites/christinecomaford/2016/06/25/how-leaders-bust-unconscious-biases-in-business/#671068252c66 Forbes.com 6/25/16