If you had to guess what percentage of today’s AI change initiatives fail to reach production — meaning, the organization invests in a platform or system but never gets to from pilot to rollout — what number would you guess? Maybe 40%? 50%?
The true number is a staggering 95%. A tiny 5% of investments in AI are currently producing anything of value for the organizations making those investments. It’s no big secret why. Change initiatives of any kind are notoriously hard. McKinsey estimates 74% of general change efforts fail. Introduce to that equation AI — a technology that represents a level of change and uncertainty most people have never experienced before— and people are likely to throw their hands up and just get back to business and usual.
Organizations that want to ensure their AI transformation moves the culture somewhere should focus more on the human side of change. Even according to the CHRO of tech company IBM, working out the technology for widespread AI transformation is maybe 15% of the challenge. The rest is a deeply human challenge. As cognitive scientists who have studied change at scale for decades, we believe we have an approach that can be far more effective. It all comes down to starting with how the brain works.
AI Change Is a Human Problem
The reason AI transformation is so hard has to do with how our brains are wired to respond to threats and rewards in our environment. The SCARF® Model captures five domains that the brain is highly sensitive to: status (our reputation), certainty (our ability to make predictions), autonomy (our sense of control), relatedness (our belonging in the group), and fairness (our desire for equitable outcomes).
Daily life triggers many of these threats on a small scale from time to time. AI transformation, however, poses strong risks to all five, all at once. We’re worried it will take our jobs, which makes us feel both unimportant, and uncertain for our future. We’re resistant to giving up control, feeling less needed by the group, and because nobody consulted us, the whole thing may feel completely unfair. The result is a five-alarm fire of SCARF that sends our threat-sensitive brains into a negative spiral. At this level of threat, no good thinking can take place. We become panicky, anxious, and overwhelmed.
Notice how this threat state is entirely cognitive in nature. The tech itself isn’t standing in the way of changing an organization’s culture or practices. Granted, using AI well does demand a bit more effort from employees, but the bulk of the change must happen between workers’ ears. For organizations to change with AI, leaders must focus on the human, even more than the tech those humans are using.
The bad news is, our recent survey data paint a grim picture: Just 5-30% of employees are “AI fluent” — that is, they partner effectively with AI to directly enhance their thinking. The remaining workforce either doesn’t use AI at all, or they mostly offload their work to save time and energy, even if the results aren’t as good. That leaves a 70-95% opportunity gap in most organizations to upskill their teams to start getting the most out of AI.
So, what exactly can leaders do to close the gap?
Change Management At Scale
Neuroscience suggests a framework for culture change that NLI has put to use in organizations of all sizes and industries: Priorities, Habits, and Systems, or PHS. Organizations must articulate their priorities for the change, define the habits that will embody those priorities, and systematize those habits so they stick and become the default. Every kind of culture change can map to PHS, including AI transformation.
To begin with, a set of priorities for why people need to be AI fluent should involve offsetting SCARF threats with SCARF rewards. For example, if leaders know their teams feel an autonomy threat from AI, they can prioritize autonomy in other aspects of using AI, such as letting people get comfortable with it on their own time, rather than rushing them into mandated use. Or if a leader senses fairness threats, they can clarify the purpose behind the initiative so people are clear on the “Why” of the rollout. Even if they don’t agree with the change, they’ll at least understand why it’s happening, which can calm a threatened brain.
With habits, teams should follow the science and focus on building one habit at a time, over time, with the whole organization building those habits together. This approach leverages the power of social learning to get people speaking the same language and moving in the same direction. As it pertains to AI transformation, cognitive science can help us identify the key habits that help people become more AI fluent. While this will be a little different depending on the use of AI at an organization, generally the habits people need are in the category of greater metacognition - or thinking about thinking. Whether it’s more flexible thinking, or more vigilance, people need a stronger set of skills for observing their own and others' thought processes themselves. One of the best ways to do this is to increase people’s understanding of everyday brain processes, or what we call ‘neuro intelligence’.
Finally, once you have the right set of priorities and habits, the right systems will include people having easy access to the right tools when they need to use them. Any friction between a person’s desire to practice the habit and actually taking action will disrupt the overall change effort. Systems should make habits easy to practice — in this case, enabling AI fluency.
NLI has decades of experience in driving large scale change initiatives. We believe that the right PHS strategy can bring over 75% of employees to AI fluency, at any scale, in under a quarter. AI coaches such as NLI’s NILES can be powerful tools for lifting AI Fluency, one habit at a time, at any scale, in any language, fast. The key is not just giving people access to an AI coach, but creating an integrated campaign that is both shared to everyone at once, and also deeply personalized, that takes people on a week by week journey of trying new things. This, brain science suggests, is how we move employees from feeling afraid of a technology they don’t understand, to feeling excited to explore its possibilities.
To learn more, listen to our recent podcast on this topic here, or join Dr Rock as he discusses how to build your AI fluency roadmap. Register here.
.avif)
