AI is unlikely to replace fields like radiology because of Jevons Paradox. By making scans cheaper and faster, AI increases the overall demand for scans, which in turn can increase the total number of jobs for human radiologists to manage the higher volume and complex cases.

Hastings points to radiology as a case study for AI's counterintuitive economic effects. While AI is superior at image processing, it didn't eliminate jobs. Instead, it made MRIs cheaper, leading to more scans and a *shortage* of radiologists needed to approve AI findings.

AI's most significant impact won't be on broad population health management, but as a diagnostic and decision-support assistant for physicians. By analyzing an individual patient's risks and co-morbidities, AI can empower doctors to make better, earlier diagnoses, addressing the core problem of physicians lacking time for deep patient analysis.

AI tools like ambient scribing are preventing physician and nurse burnout by automating administrative tasks and saving hours each day. This serves as a critical retention tool for a system facing a massive labor shortage, allowing experienced professionals to stay in their jobs longer.

The most effective AI strategy focuses on 'micro workflows'—small, discrete tasks like summarizing patient data. By optimizing these countless small steps, AI can make decision-makers 'a hundred-fold more productive,' delivering massive cumulative value without relying on a single, high-risk autonomous solution.

An effective AI strategy in healthcare is not limited to consumer-facing assistants. A critical focus is building tools to augment the clinicians themselves. An AI 'assistant' for doctors to surface information and guide decisions scales expertise and improves care quality from the inside out.

The most tangible ROI for AI in healthcare today isn't in complex diagnostics, but in operational efficiency. AI scribes that free up doctors, intelligent call centers that triage patients correctly, and automated claim management are solving major bottlenecks and fighting burnout right now.

A traditional IT investment ROI model misses the true value of AI in pharma. A proper methodology must account for operational efficiencies (e.g., time saved in clinical trials, where each day costs millions) and intangible benefits like improved data quality, competitive advantage, and institutional learning.

Many radiology AI tools aim to improve disease detection, but radiologists can already do this incredibly fast. The real bottleneck is the cognitive load of synthesizing findings from thousands of images into a report tailored for a specific referring clinician. AI should target this communication and workflow challenge to reduce burnout and save time.