Linking To And Excerpting From European Journal of Radiology Artificial Intelligence’s “Perspective: AI productivity will not benefit employed radiologists”

Today, I review, link to, and excerpt from the European Journal of Radiology Artificial Intelligence‘s “Perspective: AI productivity will not benefit employed radiologists”. [PubMed Abstract] [Full-Text HTML] [Full-Text PDF]. Volume 3, September 2025, 100033

All that follows is from the above resource.

Highlights

• Radiology is the main focus of medical AI, yet few debates focus on who benefits.
• AI raises imaging output which could reduce the value of radiologists’ labour.
• Most productivity gains will go to employers, vendors, and private-equity firms.
• History shows automation boosts efficiency while reducing labour’s share of income.
• As AI redefines roles, radiologists should seek equity, specialise, or pivot.

Abstract

Debates about AI in radiology typically ask whether it will augment or replace radiologists. It is less common to ask who profits from improved productivity. AI systems already interpret high-volume studies, such as screening mammograms, at expert-level accuracy: a recent Swedish trial showed AI safely reduced radiologist workloads by 44 %. Economist James Bessen shows that automation tends to shift value from labour to capital. Following Bessen, we predict that the potential labour savings of AI will primarily benefit employers, investors, and AI vendors, not salaried radiologists. Radiologists should be aware of this trend and where appropriate adopt strategies to navigate AI disruption, such as gaining equity in their practice, specialising in areas resistant to automation, or transitioning to alternative career paths.

Keywords

Artificial intelligence; Radiology; Machine learning; Diagnostic imaging; Mammography screening; Labour productivity; Health economics; Automation

From typists to truckers, people have long worried whether automation will complement or disrupt their work. This debate divides radiology into two camps. The augmentation camp believes AI will handle routine tasks and free up doctors to focus on complex interpretations, interventions, and ‘soft skills’. The replacement camp believes that AI will soon interpret diagnostic imaging to an above-human standard and make many radiologists redundant. The camps can coexist: AI may boost some radiologists and displace others. Yet, most commentators overlook an uncomfortable question: as AI becomes more capable, what happens to the value of radiologists’ work? In other words, cui bono – who benefits from the productivity gains?

Medical AI researcher Dr Curtis Langlotz coined the augmentation camp’s best-known credo: “AI won’t replace radiologists, but radiologists who use AI will replace those who don’t.” In his RSNA 2017 keynote, Langlotz framed AI as an “autopilot for radiologists” that will reduce burnout and tackle backlogs without compromising care [1]. His message countered a climate of fear triggered the previous year by computer scientist Geoffrey Hinton, often called the ‘Godfather of AI’. Hinton gave a vivid image to the replacement camp when he described radiologists as “like the coyote that’s already over the edge of the cliff but hasn’t yet looked down, so doesn’t realise there’s no ground beneath him.” To rub salt in the wound, he added, “we have plenty of radiologists already” and “people should stop training radiologists now.” [2]. Hinton’s remarks worried many. In 2018, 44 % of US medical students said AI made radiology less appealing [3].

Despite fears of obsolescence and partly reassured by optimists such as Langlotz, US radiology residency applications in 2019 reached a nine-year high [4]. Radiology careers remain popular because putting AI into clinical practice is unexpectedly difficult. Tools cost too much [5], integrate poorly [6], and are not trusted by doctors or patients [7]. Hinton’s 2016 prediction that AI would outperform radiologists within five years [2] was wrong. In 2021, 99 % of tools still required human oversight [8].

Yet for AI firms and investors, radiology remains too good an opportunity to ignore. It offers what they value most: a large market with high-quality datasets. As such, radiology is by far the top target for medical AI. By December 2024, 76 % of all AI-enabled medical applications cleared by the FDA target radiology [9]. Meanwhile, ageing populations, rising chronic disease, and cheaper imaging have driven up imaging volumes. Too few radiologists exist to meet demand: the UK expects a 40 % shortage of consultants by 2028 [10], and the US a gap of 122,000 by 2032 [11]. In 2025, Hinton appeared to repent, telling The New York Times he spoke “too broadly” and was referring only to image analysis [12]. He now says AI will make radiologists more efficient and accurate – a measured position leaving him room to be right either way. Again, this sidesteps the question: what will these changes mean for radiologists’ careers?

In two recent LinkedIn articles [13], [14], we argued that AI-driven productivity gains in radiology will primarily benefit vendors and employers, not employed radiologists. These articles received 90 comments from doctors and entrepreneurs divided between optimism, anxiety, and uncertainty. Once AI matches radiologists’ reporting accuracy, practices can reduce their largest expense: salaries [15]. For instance, Sweden’s national breast cancer screening programme showed that AI could safely replace one of two human readers, cutting workload by 44 % [16]. This means one radiologist plus AI can now do the job of two. Our core argument is that once AI surpasses human accuracy, fewer radiologists will be employed. These staffing decisions are increasingly not made by radiologists, but department heads, practice owners, and private equity firms. In US practices, for example, between 2012 and 2024 radiologist ownership fell from 63 % to 46 % [17], while private equity ownership rose from 1 % to 13 % [18].

Among the 90 replies, one of the most insightful came from radiologist and health-tech investor Dr Amine Korchi: “technology doesn’t kill professions; it reshapes them.” [13]. Korchi is right, automation rarely eliminates fields entirely. Yet economist James Bessen, who helped launch desktop publishing in the 1980s, demonstrates that such reshaping often leads to unequal rewards. When software automated hot-metal typesetting, employment for graphic designers grew more than fourfold (1979–2007), but these new roles required “considerable” reskilling, median pay stagnated, and the wage advantage from experience diminished as “technology and organisation of work… [was] in constant flux.” [19] (p112). Most productivity gains thus flowed to employers and a few rapid self-learners – an outcome radiology now risks repeating.

Many radiologists assume workforce shortages will protect them from automation. But history suggests the opposite: automation is most profitable when labour is scarce and expensive. High radiologist salaries increase the pressure to automate. Once adopted, automation tends to reduce labour’s share of the value it creates. As economists Acemoglu and Restrepo argue, automation “increases the size of the pie, but labour gets a smaller slice.” [20]. This displacement effect, where capital substitutes for human tasks, is why productivity gains come at labour’s expense. For instance, in Bessen’s study of the US textile industry he charted how a weaver in 1900 could produce 50 times more cloth than a one in 1800. Rising productivity made cloth cheaper and demand for it soared, as consumers discovered the joy of owning more than one outfit. This created four times more weaving jobs. However, people have a finite desire for cloth. Demand at some point can no longer keep up with productivity, and a tipping point arrives. After 1920, output per worker kept rising by about 3 % a year, but demand for cloth stagnated and employers cut jobs [19] (pp84–100). Bessen’s inverted U-curve charts this boom-and-bust cycle for labour: on the way up, lower costs open markets and create jobs. On the way down, machines replace labour, and the job market shrinks.

Could radiology be approaching its own downward slope? Past innovations like CT, PACS, and voice recognition boosted productivity, yet never enough to exceed demand. Today, AI systems increasingly perform at or above radiologist level. Consider three examples from the past year: in breast cancer screening, a South Korean trial shows that AI-assisted readers detect 13.8 % more cancers than radiologists working alone [21]. In MSK radiographs, a Finnish emergency-department study finds that two AI algorithms match expert MSK radiologists at 89 % accuracy [22]. In chest X-ray reporting, a DeepMind study shows that 75 % of radiologists judge AI-generated reports preferable or equivalent to expert-written ones [23]. Labour is scarce today, but if AI tools can increase productivity by orders of magnitude, practices will employ fewer radiologists (see Fig. 1). As AI capabilities grow, radiologists will shift from reading to reviewing, from interpreting to confirming. Tasks such as biopsies and patient consultations will likely remain in human hands. Radiology will continue to exist, but the nature and value of the work will change. For some, this brings opportunities. For others, a loss of autonomy, status, or income.

Fig. 1. Applying the Bessen U-curve to radiology shows how productivity gains can raise jobs before cutting them.

Radiologists cannot control the pace of AI, but they can prepare a clear ‘Plan B′. Some may seek equity in a practice, move into education or industry, or shift toward procedures and subspecialties that automation is less likely to affect. Younger radiologists should think not only about how to adapt, but how to build additional income streams. Those later in their careers may focus on securing roles that are harder to replace or on reducing clinical time on their own terms. Langlotz’s idea – that radiologists who use AI will replace those who do not – is partly true. But for employed radiologists, the more likely outcome is that their profession will be less in demand and less well paid. Those who remain in the field will do different work, with no guarantee that employers will value, reward, or respect it as they once did.

Employers, private equity firms, and AI vendors will capture most of the profit from AI-driven productivity. Employed radiologists are unlikely to receive these gains. Without structural change in how value is shared, increased productivity will come at the expense of those who still work.