I recently leased a Tesla with Full Self-Driving software. I had the feature in my previous car and never used it. This time, I expected to leave it alone again.
Then my 22-year-old son, Max, took the initiative and activated it. During the roughly hour-long trip home, I grew more comfortable with the system. I now estimate that I use it for more than 85 percent of my driving.
Access to the technology was settled. My willingness to use it was not. The intervention that changed my behavior came from someone I trusted, outside the company’s control.
Max was not a target receiving a message. He was an actor changing someone else’s decision. A profile of his age, income and interests would have revealed little about his power to alter my experience of the car.
That poses a problem for businesses. The people they want to reach can influence other buyers, produce their own interpretation of a brand, organize support or opposition and change the conditions in which a company operates. Artificial intelligence expands those capabilities and reduces the time and effort needed to exercise some of them.
Yet a company can acquire the newest tools and retain an inadequate model of the person: someone to reach, persuade and convert. The gap opens when targeting gets more precise while the relationship remains unchanged.
The change is in power
The explanation that people have more information is decades old. It falls short. What matters is what they can do with that information and with the relationships they have built.
Research distinguished those sources of power before the current expansion of AI. In Consumer Power Evolution in the Digital Age, Lauren Labrecque and her co-authors identified power based on demand, information, networks and collective action in 2013. [1]
Power builds through conditions that segmentation barely captures. A person has relationships that give them access to others and credibility that can carry more weight than an advertisement. They can document an experience, distribute it, gather support and turn an individual decision into a reference point for many more. They buy, work, invest and influence through relationships that cross companies’ internal boundaries.
AI adds the ability to produce and execute. It helps people research, compare proposals, build arguments and create content with fewer of the resources those tasks once required. Networks add distribution speed. What someone knows, creates or recommends can enter other people’s decisions while an organization is still considering how to respond.
Together, those conditions make someone an actor with power over business outcomes. AI expands their reach: the empowered actor becomes a super-empowered actor. Their influence depends on relationships, credibility and context, but the means of exercising it are multiplying.
For a company, the consequences begin before the campaign. They change which product is worth developing, what experience can sustain a promise and which internal behaviors build trust. Marketing, reputation and culture help leaders understand and act on those conditions. That is the basis for their place in major business decisions: they influence whether a business bet can secure adoption, support and staying power.
The idea connects with superagency, Reid Hoffman and Greg Beato’s concept, which McKinsey has taken up to explore how AI expands human capabilities. The evolution toward a super-empowered actor offers a way to understand the consequences for relationships between people and companies. [2]
Power remains unequal. Platforms control distribution and data; companies retain resources individuals lack. But a person has more means of intervening, without becoming immune to the influence of others.
The evidence reaches decisions
An experiment with 776 Procter & Gamble professionals found that an individual assisted by AI could match the performance of a two-person team without it on the innovation tasks studied. Individual capability can already change how work gets done. [3]
It can also shape preferences. Experiments published in Nature in 2025 found that conversations with AI advocating for a candidate changed participants’ stated preferences. The result measures attitudes rather than actual votes, but shows the tools’ potential to persuade. [4]
Trust still determines which influence takes hold. Research on technology acceptance and trust in automation helps explain why access to a tool does not ensure its use: perceived usefulness, ease of use and trust aligned with its capabilities matter. Max had credibility with me. Experience did the rest. [5]
That relationship also has commercial consequences. Yale researchers estimated that, without the effect of Elon Musk’s partisan activities, Tesla would have sold at least a million additional vehicles in the U.S. between October 2022 and April 2025. It is a counterfactual estimate, subject to the study’s assumptions. The boardroom question is direct: how much demand depends on what a brand represents, as well as what its product does? [6]
What are we waiting for?
People’s capabilities are advancing on their own timetable. Faster message production does little if the mindset for understanding their power stays the same. A customer can influence a community while the company is still deciding which department should respond.
That is why marketing, reputation and culture should have a seat at the table where leaders decide which product to launch, how to introduce AI or how to transform the organization. Trust, adoption and the ability to mobilize people help determine whether those bets succeed. Those economic stakes make the concerns we still call soft into hard business.
Approaching them as a system is a practical consequence. At a bank, the promise of an AI-enabled service, the customer experience and employee incentives jointly produce trust or friction. Planning for that relationship and tracking it through a dashboard tied to business outcomes allows decisions to be adjusted before a difficulty becomes a crisis.
Max changed my relationship with technology I already had. AI is now expanding what people like him can do with their relationships, initiative and credibility. Understanding that power should change both who participates in decisions and how those decisions are made.
Expecting an older mindset to create that connection would be like counting on Max, at 22, to become enthusiastic about Full Self-Driving and persuade me to use it after reading a 400-page manual tucked into an elegant leather case in the glove compartment. The manual could be flawless. The world it was written for has moved on.
Sources
- 1. Labrecque, L. I. et al. (2013). Consumer Power Evolution in the Digital Age. Journal of Interactive Marketing, 27(4), 257–269. View source
- 2. McKinsey (2025). Superagency in the workplace Empowering people to unlock AI’s full potential. Discusses Reid Hoffman and Greg Beato’s concept. Super-empowered actor is the author’s proposed synthesis, rather than a category validated by the report. View source
- 3. Dell’Acqua, F. et al. (2025). The Cybernetic Teammate A Field Experiment on Generative AI Reshaping Teamwork and Expertise. NBER Working Paper 33641. Findings concern the tasks studied. View source View source
- 4. Lin, H. et al. (2025). Persuading voters using human–artificial intelligence dialogues. Nature, 648, 394–401. Experiments on stated preferences. View source
- 5. Davis, F. D. (1989). Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology. MIS Quarterly, 13(3), 319–340. Lee, J. D. and See, K. A. (2004). Trust in Automation Designing for Appropriate Reliance. Human Factors, 46(1), 50–80. The connection to the personal account is the author’s interpretation. View source View source
- 6. Gillingham, K. et al. (2025). The Musk Partisan Effect on Tesla Sales. NBER Working Paper 34413. Counterfactual estimate, rather than a count of canceled orders. View source
