Article by Shana Lynch: “…After years of fast expansion and billion-dollar bets, 2026 may mark the moment artificial intelligence confronts its actual utility. In their predictions for the next year, Stanford faculty across computer science, medicine, law, and economics converge on a striking theme: The era of AI evangelism is giving way to an era of AI evaluation. Whether it’s standardized benchmarks for legal reasoning, real-time dashboards tracking labor displacement, or clinical frameworks for vetting the flood of medical AI startups, the coming year demands rigor over hype. The question is no longer “Can AI do this?” but “How well, at what cost, and for whom?”
Learn more about what Stanford HAI faculty expect in the new year…As the buzz around the use of GenAI builds, the creators of the technologies will get frustrated with the long decision cycles at health systems and begin going directly to the user in the form of applications that are made available for “free” to end users. Consider, for example, efforts such as literature summaries by OpenEvidence and on-demand answers to clinical questions by AtroposHealth.
On the technology side, we will see a rise in generative transformers that have the potential to forecast diagnoses, treatment response, or disease progression without needing any task-specific labels.
Given this rise in available solutions, the need for patients to know the basis on which AI “help” is being provided will become crucial (see my prior commentary on this). The ability for researchers to keep up with technology developments via good benchmarking will be stretched thin, even if it is widely recognized to be important. And we will see a rise in solutions that empower patients to have agency in their own care (e.g., this example involving cancer treatment)…(More)”.
