Nick Lichtenberg expected curiosity when he spoke openly about his heavy use of artificial intelligence in journalism. What he received instead was backlash. Editors took notice with interest. Reporters recoiled. Online critics labeled him lazy. Some journalists admitted in private they followed the same practices yet refused to say so aloud. One reader demanded a coffee meeting to explain exactly why he had it all wrong.
This reaction ran deeper than ethics in media. It touched something personal. Something ancient. Lichtenberg, business editor at Fortune, found the clearest explanation not from fellow reporters but from neuroscientist Vivienne Ming.
Ming has spent three decades blending AI with human cognition. Her early work included a facial analysis system funded partly by the CIA. She built learning algorithms into cochlear implants. Her companies have tackled hiring bias, Alzheimer’s detection, and postpartum depression prediction. When she speaks about AI, she draws from long experience wiring machines directly into minds.
Last year Ming ran a telling experiment at UC Berkeley. She asked students to forecast real-world events on Polymarket, a platform where participants risk actual money on outcomes ranging from commodity prices to geopolitical shifts. The questions defied rote knowledge. No amount of cramming would reveal the future price of oil six months out.
She monitored some participants with EEG devices. The brain scans delivered a shock. Most students, labeled automators by Ming, simply passed queries to the AI and copied the output. Their gamma wave activity — a marker of deep cognitive effort — fell by about 40 percent. The drop matched the difference between solving a tough math problem and watching television. These were top students at a leading university. With powerful AI at hand they had turned into expensive copy-paste operators.
A second group, the validators, fared even worse. They used AI only to confirm existing beliefs. They ignored contradictory data. Their performance sank below that of the AI working solo.
But a small cohort behaved differently. Ming estimates they represent 5 to 10 percent of people. Their exchanges with AI showed no clear leader. Human and machine challenged each other in loops. The person offered hunches and pushed into uncertain territory. The model anchored them with evidence and corrected overconfidence. Both sides updated in turn. Ming calls them cyborgs. They beat the strongest individual humans. They beat the best standalone AI systems. Their results rivaled those of Polymarket’s expert traders who bet with real capital.
The model size hardly mattered. State-of-the-art systems performed no better than cheaper open-source versions for these users. What separated winners was the human. Ming identified four traits that predicted cyborg success with striking consistency: curiosity, fluid intelligence, intellectual humility, and perspective-taking. These same qualities, when measured in children, forecast lifetime earnings and even mortality risk. They shape how people meet the world.
Curiosity keeps the search alive after the first decent answer appears. Fluid intelligence handles problems without familiar templates. Intellectual humility allows belief revision when the machine pushes back. Perspective-taking models other viewpoints and surfaces options data alone might miss.
Kate Smaje reached similar conclusions from a different vantage. As McKinsey’s global leader for technology and AI she has advised clients across industries and continents. When asked which human abilities remain essential, she lists judgment, conceptual problem-solving, empathy, and trust. The overlap with Ming’s traits is obvious. Judgment aligns with fluid intelligence. Conceptual problem-solving echoes curiosity. Empathy matches perspective-taking. Trust reflects intellectual humility.
Organizations stumble not because they lack technology, Smaje told Fortune. They falter on the human changes required to surround it. “I fundamentally believe that the world is going to need really great humans,” she said.
Bret Greenstein, chief AI officer at West Monroe, sees the same resistance in client engagements. People tie their identity to specific tasks — the one who builds the PowerPoint, fills the spreadsheet, writes the report. AI threatens that sense of self. It forces a shift from doing things to making decisions. Greenstein’s team calculated that AI delivered output equal to 320 full-time employees in six months without new hires. Some employees lit up at the possibilities. Others shut down.
This pattern repeats across sectors. A Udacity survey of executives, managers, and employees found only 9 percent would replace their entire workforce with AI tools. Seventy percent of respondents held management roles. Most prefer working with humans who bring institutional knowledge, creativity, critical thinking, and accountability. Victoria Papalian, COO at Udacity, noted the push to substitute workers with AI agents meets resistance from managers as well as staff. The CIO report on that survey appeared in March 2026.
Productivity data tells a complicated story. The Microsoft AI Diffusion Report for the first quarter of 2026 showed global AI adoption rising 1.5 percentage points in three months. Goldman Sachs economists reported that workers who master the tools save 40 to 60 minutes daily. Yet broader surveys paint a different picture. A WalkMe study of 3,750 executives and employees across 14 countries found 54 percent had bypassed company AI tools in the past month to finish work manually. Another 33 percent had not used them at all. Roughly 80 percent of enterprise workers either avoid or reject the technology their employers fund heavily.
Those employers spent more on digital transformation — budgets rose 38 percent year-over-year to an average $54.2 million — yet 40 percent of that investment underperformed due to adoption failures. Workers lose the equivalent of 51 days per year to technology friction, a 42 percent increase from the prior year. The gains for skilled users appear offset by losses among everyone else. Steve Hanke, economist at Johns Hopkins, told Fortune the productivity surge many predicted has not materialized. “AI didn’t deliver,” he said.
Public sentiment has cooled. The New York Times reported in February 2026 that AI enthusiasm lags behind earlier technology booms. More than a third of Americans in one poll feared the technology could end human life. Most said they would not pay extra for AI features on their devices. Eighty percent of firms in a National Bureau of Economic Research survey reported no measurable effect on productivity or employment. Sam Altman of OpenAI admitted greater resistance to adoption than he anticipated.
Backlash appears in multiple forms. Some target data centers required to power AI systems. The New York Times podcast Hard Fork discussed in April 2026 how opposition has turned political and occasionally violent. Maine passed a temporary moratorium on large data centers. Protests have arisen in multiple states against new facilities. Steve Bannon and Bernie Sanders, rarely aligned, both warn that AI threatens the working class. The Atlantic explored in May 2026 how this bipartisan anxiety could grow uglier if job losses accelerate.
Grassroots resistance also surfaces in creative fields and daily life. Computer Weekly described in February 2026 an emerging “analog living” movement that rejects constant efficiency in favor of creativity and community. Some workers quietly rebel by ignoring AI mandates. Others produce what critics call workslop — low-quality output generated quickly then passed along. A Guardian investigation in April 2026 found 40 percent of workers encountered such material monthly and spent hours correcting it. Executives often claim productivity gains while employees report feeling busier and less effective.
Vivienne Ming warns that many fears about AI mask deeper worries about other people. A colleague who masters these tools doesn’t just work faster. That person changes the definition of the job itself. The rest feel indicted by the new standard. Ming calls this a cognitive divide. In her view most interesting problems are ill-posed — open-ended, without clear templates. Society built education and hiring systems around well-posed questions that machines now answer better.
Her forthcoming book Robot-Proof argues the solution lies in building better people. Not through narrow skills training that AI will soon master but through the human capacities that grow more valuable as machines advance. Curiosity. Adaptability. Humility. The willingness to engage in genuine back-and-forth with intelligent systems rather than treat them as oracles or clerks.
Recent research supports parts of this view. A Harvard Business School study found AI can replicate some benefits of teamwork for individuals, improving output quality and saving time. Yet the highest performance still came from full human teams combined with AI. Fabrizio Dell’Acqua, one of the researchers, advised companies to treat the technology as an enhancer of collaboration rather than a replacement for it.
Leaders face a choice. They can chase narrow automation that yields short-term gains but breeds resentment, friction, and mediocre results. Or they can redesign roles, workflows, and incentives around the traits that let humans and machines reach superhuman outcomes together. The latter path demands more than technology investment. It requires cultural shifts, new forms of training, and honest reckoning with what work means when machines handle routine cognition.
Greenstein at West Monroe sees both reactions in organizations. Some teams embrace the change and discover new levels of output. Others cling to old identities and fall behind. The gap between those groups is widening. Ming’s EEG data suggests the difference appears at the level of brain activity.
The stakes extend beyond individual careers. Companies that fail to develop these human strengths risk stalled adoption, wasted spending, and internal division. Economies that cannot bridge the cognitive divide may watch productivity promises evaporate into friction and backlash. And workers who treat AI as a crutch rather than a partner may find themselves on the wrong side of a divide that predicts not just professional success but broader life outcomes.
Lichtenberg continues his experiments with AI in reporting. He uses it to synthesize interviews, spot connections, and draft early versions. Then he checks every claim, challenges every assumption, and decides what the story truly demands. The process feels like an ongoing dialogue rather than delegation. It has shortened some tasks dramatically. But it has not removed the need for judgment, skepticism, or the distinctly human ability to sense what matters.
Most people, Ming’s data suggests, will not follow that path naturally. They will need deliberate cultivation of the qualities that turn AI from a replacement into a genuine collaborator. Organizations that recognize this reality early and act on it may capture gains that remain elusive for the majority. Those that ignore it will watch their expensive tools gather digital dust while resentment grows.
The future belongs to the cyborgs. Not because machines have won. But because a small share of humans have learned to meet them as partners.
WebProNews is an iEntry Publication