The Wall Street Journal recently published the headline: “Walmart CEO Issues Wake-Up Call: ‘AI Is Going to Change Literally Every Job.’” Read the article here. At first glance, it reads like a bold declaration from the country’s largest private employer. In reality, it is another example of corporate hype about artificial intelligence, full of sweeping promises, recycled narratives, and very little evidence.
We have seen this script before. Over a decade ago, IBM announced that its Watson platform would revolutionize healthcare and business. Watson was supposed to diagnose cancer, transform finance, and redesign industries. Billions were invested, but the project collapsed under the weight of its own hype. Watson Health was quietly dismantled and sold off, remembered today more as a cautionary tale than a success.
Doug McMillon is now repeating the same rhetorical playbook, insisting that AI will change “literally every job.” The comparison is telling. When large corporations oversell technology, reality has a way of catching up.
What evidence supports these new claims? None. The company cannot point to specific roles eliminated or transformed by AI. It cannot show measurable productivity gains, significant cost savings, or improved customer service directly attributable to the technology. Instead, it offers vague promises. The size of the workforce will remain “flat.” The “mix of jobs” will evolve. Training will be needed. These are not results. They are placeholders, bold words without proof behind them.
As McMillon told the Journal, “Maybe there’s a job in the world that AI won’t change, but I haven’t thought of it.” This is not analysis. It is exaggeration without evidence.
The company’s own history with automation shows why skepticism is justified. In 2017, Walmart began testing shelf scanning robots built by Bossa Nova Robotics. By 2018, the machines were in hundreds of stores. In 2019, Walmart announced plans to expand them to 1,000 locations. By late 2020, the project was over. The contract was canceled, the robots were pulled from stores, and Bossa Nova laid off staff.
The explanation was straightforward. Employees already stocking shelves and fulfilling online orders could check inventory just as effectively. The robots were costly, impractical, and unnecessary. Walmart had invested millions in a highly publicized project, only to retreat and return to human labor. That experience should have been a lesson in humility. Instead, the company has returned with another sweeping narrative about AI.
The idea that artificial intelligence will transform every job is more fantasy than fact. AI is not an autonomous mind capable of replacing humans across the board. It is a collection of statistical models trained on massive datasets. These systems are useful for narrow tasks such as forecasting demand, classifying products, or generating text. They are also brittle, limited, and prone to errors outside their training scope.
AI will not change plumbers fixing leaks, bakers kneading dough, truck drivers navigating icy roads, or cashiers helping elderly customers. At best, it will affect certain clerical and repetitive tasks. To suggest otherwise is fearmongering disguised as foresight.
It is also worth remembering who is making these claims. Doug McMillon is a retail executive, not an AI researcher. His expertise is in merchandising and management, not machine learning. When actual AI scientists speak, they emphasize limits such as hallucinations, bias, and the need for constant supervision. These warnings are absent in the corporate narrative. Instead, the microphone is handed to executives whose incentive is to exaggerate.
Even the examples highlighted sound less like breakthroughs and more like rebranded tools. The so called “agents” are simply chatbots, and the new “agent builders” are employees tasked with making more of them. Chatbots have existed for decades. Calling them agents does not make them revolutionary. This is digitization, not transformation.
The pledge to keep the workforce “flat” at 2.1 million over the next three years is equally misleading. A flat headcount is not the same as job security. A company can maintain the same number of employees while turning more jobs into part time positions, increasing turnover, or destabilizing schedules. The announcement comforts workers on the surface while signaling to investors that automation will control labor costs. It is spin designed to serve both sides, not a guarantee of stability.
The human cost of these experiments is never mentioned. When shelf scanning robots failed, workers had to pick up the slack. When AI scheduling systems are introduced, employees often find their hours unpredictable. These are workers already struggling with low wages and limited career growth. If the company were truly committed to resilience, it would invest in stability and fair pay, not conferences about AI.
This pattern extends beyond one retailer. Executives at Ford, JPMorgan, and Amazon repeat the same script. The strategy is clear. By warning of technological disruption, they discipline workers and excite investors. Workers, fearful of being replaced, are less likely to demand higher wages. Investors, enticed by visions of efficiency, reward the stock. The story of AI “changing every job” serves management’s interests far more than it reflects technological reality.
The truth is that AI is not magic. It is statistics on a massive scale. Language models like ChatGPT predict the next word. Demand systems predict sales spikes. Image generators predict pixels. These systems can be powerful in narrow contexts, but they lack judgment, imagination, and empathy. Those are the qualities that make human work meaningful. Pretending otherwise is not just misleading. It is manipulative.
The Journal’s framing of these remarks as a “wake up call” misses the point. The real wake up call is not that AI is coming to change every job. It is that we have heard this all before. From IBM Watson to failed robotics projects, the cycle repeats: overpromise, headline, quiet retreat. AI will not transform every job. It will not erase workforces. It will not create a science fiction vision of robot run retail. At best, it will provide incremental tools for narrow tasks. At worst, it will destabilize already precarious jobs.
The public deserves better than recycled hype. Workers deserve clarity, not fear. Investors deserve facts, not marketing spin. Until corporations can back their AI claims with evidence, we should treat these statements for what they are: noise designed to impress, not truth designed to inform.
