The rapid rise of artificial intelligence has sparked a wave of bold proclamations that machines will soon take over human jobs, a narrative that grabs headlines but often misses the nuanced reality on the ground. Kevin Indig’s recent Growth Memo cuts through the hype by labeling this approach “substitution positioning,” arguing that framing AI as a wholesale replacement for people damages the very credibility that technology providers need to succeed. His analysis is noteworthy because it comes from a source known for data‑driven, cautious commentary, making his warning a signal that the industry’s storytelling may be out of step with actual labor market movements. By grounding his critique in concrete evidence from state‑level layoff disclosures and a long‑running academic study, Indig shifts the conversation from speculative fear to observable trends, offering a foundation for marketers and product leaders to reassess how they talk about AI’s role in the workplace.
Substitution positioning is essentially a modern twist on Theodore Levitt’s concept of marketing myopia, where companies define themselves by what they sell rather than the underlying customer need. When an AI vendor claims its tool will eliminate entire job categories, the implicit message to prospective buyers and current employees is that their roles are at risk. This framing triggers a defensive mindset: decision‑makers become hesitant to adopt a solution that could be perceived as a termination notice, and workers may resist or even sabotage implementation out of self‑preservation. The short‑term attention gained from sensational replacement talk is therefore offset by a longer‑term erosion of trust, making it harder for even genuinely useful AI products to gain traction in organizations that value stability and employee morale.
The discomfort surrounding these claims intensifies when the loudest predictions originate from the very firms building the technology. In early 2026, Anthropic’s CEO Dario Amodei forecast that AI would handle most or all of software engineers’ end‑to‑end tasks within six to twelve months, a timeline that has not materialized as demand for engineering talent continued to rise. Later in 2025, OpenAI’s Sam Altman suggested that AI would soon shoulder the bulk of phone‑ and computer‑based customer support, predicting a net benefit for all stakeholders. Yet shortly after those statements, hiring in customer service outpaced the broader labor market, indicating that the anticipated displacement did not occur. These examples illustrate how confident, near‑term replacement forecasts can quickly age poorly, leaving companies with a credibility gap when reality fails to match the rhetoric.
Such mismatches are not merely rhetorical missteps; they accumulate as liabilities in the perceptions of the stakeholders that AI firms depend on—buyers evaluating purchases, employees assessing job security, and regulators scrutinizing market conduct. When a company repeatedly promises imminent workforce reduction that never shows up in employment data, its future claims are viewed with skepticism, making it harder to launch new products or secure partnerships. Over time, the pattern of overpromising and underdelivering can trigger a boycott mindset among potential adopters, who may prefer vendors with more modest, evidence‑based messaging. In a competitive landscape where trust is a differentiator, the cost of broken promises can outweigh any short‑term publicity gains from sensational substitution talk.
Indig’s argument gains strength from two independent data sets that have received relatively little attention in the trade press. First, New York State enacted a regulation in March 2025 requiring firms filing mass‑layoff (WARN) notices to disclose whether technological innovation or automation played a role, and to name the specific technology if applicable. Over the following fourteen months, more than 160 companies submitted notices covering roughly twenty‑eight thousand affected workers. Notably, major players like Amazon and Goldman Sachs—both of whom have publicly discussed AI’s productivity benefits—did not check the box attributing any of those layoffs to AI or automation. This absence of attribution suggests that, at least in the documented mass‑layoff events, AI has not been identified as a direct cause of workforce reductions.
Interpreting the New York WARN data requires caution, but the pattern is striking: if AI were truly displacing workers at the scale suggested by some executive forecasts, we would expect to see at least a few companies citing it as a contributing factor in their layoff reports. The fact that none did, despite the inclusion of firms actively experimenting with AI tools, points to a more complex reality where technology may be augmenting tasks, shifting responsibilities, or creating new roles rather than outright eliminating positions. It also raises the possibility that layoffs driven by other economic pressures are being mistakenly attributed to AI in public discourse, while the actual impact of AI on headcount remains subtle or indirect.
The second data set comes from the Yale Budget Lab, which has been analyzing the Current Population Survey for over thirty‑three months to detect any measurable AI‑related displacement across the economy. By combining occupational mix metrics, industry dissimilarity measures, and AI exposure scores, the researchers attempted to isolate the effect of AI on employment levels and wage trends. Their most recent update concludes that there is no statistically or economically significant impact of AI on jobs or pay to date. In other words, the aggregate labor market shows stability rather than the upheaval that replacement‑focused narratives would predict. The Lab further notes that AI’s influence resembles the gradual, uneven adoption of earlier general‑purpose technologies like computers and the internet, which brought both augmentation and selective displacement over extended periods.
These findings do not mean AI is inert; rather, they highlight that its current integration into workflows is more akin to a tool that enhances human capabilities than a wholesale substitute. The technology tends to automate specific, repetitive components of jobs while leaving room for human judgment, creativity, and interpersonal interaction—elements that are difficult to replicate fully with existing models. This pattern mirrors past technological shifts where productivity gains emerged alongside job transformation, not outright eradication. Recognizing this nuance helps avoid the pitfall of treating AI as a binary force that either leaves work untouched or completely replaces it, a view that fuels both unwarranted optimism and unjustified fear.
The gap between what AI companies proclaim in press releases and what the employment data actually shows is what Indig describes as “AI washing”—the practice of dressing up modest productivity gains as revolutionary workforce disruption to capture attention and investment. When the public narrative promises imminent replacement but the data reveal only incremental change, skepticism builds not only among potential customers but also among talent pools that fear obsolescence. This erosion of credibility can make it harder for AI firms to recruit skilled workers, form strategic alliances, or gain regulatory approval, ultimately slowing the very innovation they seek to promote. The lesson is clear: sustainable market advantage stems from aligning marketing messages with observable outcomes rather than chasing short‑term buzz through exaggerated claims.
Understanding why the replacement frame suppresses adoption requires looking at basic human psychology. No one wants to hear that their livelihood might be rendered unnecessary by a machine, even if the statement is framed as an opportunity for greater efficiency. When an AI product is marketed with the tagline “you can do more with fewer people,” the subconscious interpretation for many is “you might be one of the fewer.” This perception triggers loss aversion, a cognitive bias where the pain of potential loss outweighs the pleasure of equivalent gain, leading decision‑makers to delay or avoid purchase. Employees who sense a threat to their roles may become silent resistors, offering minimal cooperation, or vocal opponents who actively discourage others from adopting the technology, thereby undermining implementation success.
To convert this awareness into a stronger market position, Indig proposes three concrete shifts that preserve the substance of AI’s capabilities while rebuilding trust. First, frame the value proposition around augmentation and tangible outcomes rather than elimination. Emphasize how the tool makes users more effective at tasks they already value, such as accelerating research, improving accuracy, or freeing up time for higher‑order creativity. This approach aligns with the buyer’s desire for competence and growth, positioning AI as a partner that enhances performance rather than a harbinger of job loss.
Second, replace vague, sweeping statements with precise, concrete descriptions of what the AI actually does and does not handle. Instead of claiming “AI handles the work your team used to do,” specify something like “AI performs the initial data‑gathering and summary drafting that previously consumed eight hours of analyst time each week, allowing the team to focus on client strategy and relationship building.” Such specificity prevents the audience from mentally inserting themselves into the role of the displaced worker and instead highlights a clear, limited function that complements human effort. It also makes it easier to measure ROI, as the time‑saved metric can be directly tracked and reported.
Third, exercise caution with timeline‑based predictions about when AI will replace specific job categories. The historical record shows that confident near‑term forecasts have repeatedly proven premature, tying a brand’s credibility to uncontrollable external factors. By anchoring messaging in what the technology demonstrably achieves today—supported by case studies, pilot results, or third‑party validation—companies build a reputation for reliability. When the market later observes genuine shifts, those same firms will be seen as foresighted advisors rather than alarmist forecasters, preserving trust for the long haul.
In conclusion, the evidence from state layoff disclosures and longitudinal labor surveys paints a picture of AI as an evolving augmentative force rather than an imminent job‑eliminating wave. Marketing that leans into replacement rhetoric may capture fleeting attention but ultimately damages the trust necessary for sustained adoption and industry leadership. By embracing augmentation, articulating precise capabilities, and resisting the temptation to overpromise on timelines, AI providers can differentiate themselves in a crowded marketplace, foster stronger buyer and employee relationships, and lay the groundwork for responsible innovation. The path forward is not to silence the conversation about AI’s impact, but to shape it with honesty, specificity, and a focus on mutual growth.