Artificial Intelligence and the Future of Work (2025) / Chapter Skim
Currently Skimming:
Pages 111-121
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 111... ...
: writing business and legal documents; digesting, distilling, and synthesizing research; producing presentations, charts, illustrations, and animations; performing state-of-the-art medical diagnoses and providing treatment plans; solving engineering and design problems; managing complex systems such as power grids, server clusters, and air traffic control systems; and developing educational content.69 The rapid progress of AI in these domains is illustrated in Figure 4-5, which shows that OpenAI's ChatGPT v4.0 LLM is currently able to score above the 80th percentile on numerous high school Advanced Placement exams (statistics, macroeconomics, microeconomics, and psychology) as well as on the Uniform Bar Examination, the quantitative reasoning section of the Graduate Record Examination, and the mathematics section of the SAT.70 While acing standardized tests is not equivalent to practicing successfully in a professional environment (i.e., lawyers do not take standardized tests for a living)
|
From page 113... ...
Many forms of elite expertise may become more accessible and less expensive in the AI era. These observations are consonant with a growing set of studies that seeks to classify which workplace tasks, occupations, and industries are most "exposed" to AI -- meaning that they are engaged in tasks that appear within the realm of AI's Artificial Intelligence and the Workforce 113
|
From page 114... ...
At the onset of prior technological revolutions, it was comparatively easy for contemporaries to predict that the Industrial Revolution would displace artisanal expertise -- the Luddites seem to have caught on to this idea quickly -- or that the computer era would displace worker expertise in routine clerical and production tasks. Similarly, it is self-evident that AI will substitute for some of the nonroutine cognitive labor tasks supplied by profes sional workers.
|
From page 115... ...
Rather than elite expertise becoming a gatekeeper for many professional products, it may become increasingly optional -- called for only when the stakes are high or the appropriate course of action is highly uncertain. More broadly, if the defining feature of classical computing is that it is uniquely capable of carrying out routine tasks rapidly, then the defining feature of AI is that it excels at mastering nonroutine tasks.
|
From page 117... ...
:1–44. Artificial Intelligence and the Workforce 117
|
From page 118... ...
It rests on the hypothesis that the future of expertise may borrow attributes from both elite and mass expertise. Mass Expertise Versus Elite Expertise The key distinction between mass expertise and elite expertise is the types of problems to which they are applied.
|
From page 119... ...
:3–10. Artificial Intelligence and the Workforce 119
|
From page 120... ...
When training is complete, a certified health care worker, assisted by AI, could perform a far larger set of diagnostic and care tasks than is currently feasible.85 The NP occupation provides a focal example of how a technology such as AI could be used to complement the expertise of nonelite workers. Specifically, by providing relevant frontier expertise, guidance, and a set of digital guardrails, AI could enable less expert workers to perform more expert tasks.
|
From page 121... ...
Because the stakes in patient care are high and the "correct" steps are not fully specified, expert judgment is required to draw a link from the domain of technical and procedural knowledge to any specific practical instance -- for example, evaluating whether a patient needs routine versus emergency care, determining which diagnostic tests should be admin istered, knowing when to probe for specific symptoms that the patient might not report, deciding whether or not to execute a risky life-saving procedure, and so on. Expert judgment is almost invariably gained through supervised learning, where established experts advise, coach, and (critically)
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.