- Leigh Thompson is a professor at Kellogg School of Management at Northwestern University.
- She writes that we should not fight our urges to treat AI like a person.
The pandemic was a 100-foot wave that forced managers to define and become their best "virtual self" for negotiations and other professional interactions.
Now generative AI is a mega-tsunami that will redefine leadership and work across sectors. It is simply a smarter, faster, tireless team member in many contexts.
You need to think about how the growing presence of AI at work will influence your leadership behavior and communication style, to continue performing at your best — for yourself and your organization.
I'm a social-organizational psychologist and professor at a top business school. Among other areas, I study how people's behavior changes when they engage in virtual, as opposed to in-vivo, communication. I've found that by taking some specific, proactive steps, people can dramatically improve their virtual communication intelligence, with implications for how we can optimize our interactions with AI.
In this article I offer four big-picture tips for working with generative AI, to not just techno-survive but techno-thrive in this revolutionary phase of the Digital Age, using insights and research from psychology.
1. Go with a growth mindset
Prior to ChatGPT's public launch, most managers had minimal experience using generative AI. That changed quickly as some businesses — including IBM — suggested that managers start using AI or risk losing their jobs.
The challenge is that generative AI has moved everyone to the bottom of the learning curve, forcing questions like "Do I have what it takes to make this technology work for me and my company?
The better question for leaders is, "Do I believe I can learn to leverage generative AI in a productive way?" That's because the latter represents a growth mindset.
Carol Dweck's much-cited research on fixed-versus-growth mindsets revealed that some people believe that ability is innate, genetic, or otherwise immutable. Fixed-mindset beliefs like "I'm not a spatial thinker" or "I struggle with technology" can lead to withdrawal, fear, anxiety, and avoidance of new things or challenges.
We are seeing this at an organizational level with AI, as some leaders put the brakes on using generative AI, based in part on perception of an AI-related skills gap among employees and the challenge of filling that.
That head-in-the-sand approach is best-replaced with one embodying a growth mindset: that learning and performance are driven largely by motivation, grit, and practice, , including when it comes to new technologies. Look for research-backed, practical steps to embrace a growth mindset, such as asking these two simple questions: "What can I do to help myself?" and "Is there a way to do this even better?"
2. Treat AI like a person — not a thing
Generative AI's conversational style entices users to treat the technology as human. It may be tempting to fight that. Don't.
Why? Anthropomorphizing — or treating non-human entities as humans with personalities and emotions is natural.
Some people may feel silly interacting with generative AI like a human, using personal names and such (even though we already do with Alexa, Siri, and other personal devices). But research shows we experience better motivation, learning, and performance when anthropomorphizing things in our environment.
One reason is that doing so boosts oxytocin levels in our brains, elevating trust, compliance, and team decision-making. Research shows, moreover, that attributing human qualities to one's own organization ("Our company is caring and generous") primes interpersonal emotions like empathy and accountability, with positive effects on the organization.
So doing the same with AI should have similar positive effects. That gives you the greenlight to talk to AI as you would another human, including keeping the tip below in mind.
3. Avoid power trips
By design, generative AI puts us in a position of power: we provide a clear, short prompt, and it has to react (promptly). But this commanding communication style is less common with human interactions, and certainly less advisable. So we need to understand how being in a position of power with AI affects our psychology.
To get at that I did a recent experiment with ChatGPT.
On the first day, I treated the AI like a human assistant in my prompts:
"Hi, I'd love your help identifying scientific and business publications on virtual communication." The next day I provided more terse, demanding prompts: "Give me no fewer than 10 publications on virtual communication, along with annotated bibliographies."
Initially, the responses to the two types of prompts were similar in tone and content. But as I continued using a courteous or curt tone, ChatGPT's responses matched my communication style. The platform even began to hallucinate (present non-factual information as fact) more in response to my demanding prompts, and I began to feel frustrated with the process. That's consistent with social-psychology research showing a reciprocity effect, such that people react to a given type of behavior in kind.
To go even further, my interaction with ChatGPT illuminated the danger of embracing a narrow role of any kind, as aligned with the creepy takeaway of Phil Zimbardo's infamous Stanford prison experiments — where university students assigned to be guards quickly became abusive over other students assigned to be prisoners.
The conclusion here is that to get the best of generative AI, don't play into a dictator-servant relationship. Aim to build a collaborative relationship with the technology based on courtesy and respect — just as you would, hopefully, with a human colleague.
4. Don't techno-stress
Generative AI's recent rise wounded my pride. I've always been proud of my ability to write quickly, with high quality. But ChatGPT forced me to recognize one of my core competencies has been eclipsed — big time. AI can write everything from mission statements to book proposals in a tiny fraction of the time I would have taken, with reasonable accuracy.
Like me, many are experiencing similar "techno-stress," a published term that refers to the anxiety we feel when faced with learning new technologies. And it's not just writers or programmers who are feeling the bite of generative AI.as the new technology seeps into so many areas formerly dominated by humans. But facing a shift in identity needn't be alarming or depressing. Indeed, research shows that the attitude we maintain toward identity change is key to our future happiness: perceiving an identity shift as progressive promotes greater self-worth and positive emotion.
I think of it as "techno-thriving" (my own term), or not merely adapting to new technologies but embracing them and making oneself open to learning and evolving. So how can you techno-thrive? By reassessing your strengths and weakness in service of aligning your actual and ideal selves. The idea is to go from "I'm someone who struggles with technology" to "I'm willing to learn new technologies and may be surprised by what I can do with them."
Of course, none of this is easy. As humans, we are naturally averse to change, and generative AI represents a tremendous shift in the way we work, create, and collaborate. But it's here to stay, so aim to make the most of it.
Leigh Thompson is a professor at Kellogg School of Management at Northwestern University and the author of "Negotiating the Sweet Spot: The Art of Leaving Nothing on the Table."