These 40 Jobs May Be Replaced by AI. These 40 Probably Won’t

A new study measuring the use of generative artificial intelligence in different professions has just gone public, and its main message to people working in some fields is harsh. It suggests translators, historians, text writers, sales representatives, and customer service agents might want to consider new careers as pile driver or dredge operators, railroad track layers, hardwood floor sanders, or maids — if, that is, they want to lower the threat of AI apps pushing them out of their current jobs. Why should anyone heed yet another of the myriad, sometimes conflicting reports in AI’s potential impacts on jobs? Because the researchers behind the new findings really know what they’re talking about. They all work for tech giant Microsoft, which is developing Copilot and related AI apps examined in the study. And those tools, the authors say, risk putting ticket agents and telemarketers out of work far sooner than orderlies and paving equipment operators. The Microsoft study comes as debate continues about the employment threats AI may pose to millions of people in clerical, administrative, communications, marketing, and other jobs. Executives of several tech companies, including AI developer Anthropic CEO Dario Amodei, have alternatively cheered or warned about bots automating a wide range of work tasks, potentially eliminating up to half of all white-collar and entry-level jobs in the process. Other business leaders, notably serial entrepreneur Mark Cuban, believe the tech will generate even more new positions than it erases by assuming a lot of repetitive drudge work. Microsoft’s new research doesn’t offer an opinion on the quantitative consequences on employment that AI will ultimately have. But it does provide clear indication of which 40 jobs are already using apps most frequently — and the contrasting 40 professions reflecting the tech’s lowest levels of penetration. The full text of their findings and the two rankings are available here. These 40 jobs are most threatened by AI Interpreters and Translators Historians Passenger Attendants Sales Representatives of Services Writers and Authors Customer Service Representatives CNC Tool Programmers Telephone Operators Ticket Agents and Travel Clerks Broadcast Announcers and Radio DJs Brokerage Clerks Farm and Home Management Educators Telemarketers Concierges Political Scientists News Analysts, Reporters, Journalists Mathematicians Technical Writers Proofreaders and Copy Markers Hosts and Hostesses Editors Business Teachers, Postsecondary Public Relations Specialists Demonstrators and Product Promoters Advertising Sales Agents New Accounts Clerks Statistical Assistants Counter and Rental Clerks Data Scientists Personal Financial Advisers Archivists Economics Teachers, Postsecondary Web Developers Management Analysts Geographers Models Market Research Analysts Public Safety Telecommunicators Switchboard Operators Library Science Teachers These 40 jobs are least threatened by AI Phlebotomists Nursing Assistants Hazardous Materials Removal Workers Helpers–Painters, Plasterers Embalmers Plant and System Operators, All Other Oral and Maxillofacial Surgeons Automotive Glass Installers and Repairers Ship Engineers Tire Repairers and Changers Prosthodontists Helpers–Production Workers Highway Maintenance Workers Medical Equipment Preparers Packaging and Filling Machine Operators Machine Feeders and Offbearers Dishwashers Cement Masons and Concrete Finishers Supervisors of Firefighters Industrial Truck and Tractor Operators Ophthalmic Medical Technicians Massage Therapists Surgical Assistants Tire Builders Helpers–Roofers Gas Compressor and Gas Pumping Station Operators Roofers Roustabouts, Oil and Gas Maids and Housekeeping Cleaners Paving, Surfacing, and Tamping Equipment Operators Logging Equipment Operators Motorboat Operators Orderlies Floor Sanders and Finishers Pile Driver Operators Rail-Track Laying and Maintenance Equipment Operators Foundry Mold and Coremakers Water Treatment Plant and System Operators Bridge and Lock Tenders Dredge Operators The results were obtained by analyzing 200,000 “conversations between users and Microsoft Bing Copilot.” Researchers then matched those with “measurements of task success and scope of impact, [to] compute an AI applicability score for each occupation.” The jobs with the highest use rates tended to be office positions or other work communicating data or thoughts for specific business purposes. “We find the highest AI applicability scores for knowledge work occupation groups such as computer and mathematical, and office and administrative support, as well as occupations such as sales whose work activities involve providing and communicating information,” the study says. “Additionally, we characterize the types of work activities performed most successfully, how wage and education correlate with AI applicability, and how real-world usage compares to predictions of occupational AI impact.” The upshot of that is data showing the work of ticketing agents, proofreaders, and PR specialists is already being automated at far higher levels than labor provided by housepainters and plasterers, embalmers, ship engineers, and phlebotomists — the technicians who draw blood for medical tests. But despite the study establishing a de facto ranking of the jobs threatened the most — and least — by AI, its authors ultimately waffle a bit on just how big the tech’s impact on overall employment and workplace stability might be. That’s probably not surprising, given they all work for the same Microsoft employer whose business future will largely depend on successfully developing and selling those work-automating apps to other companies. And some of the study’s disclaimers suggest underlining AI’s potential for possibly eviscerating current employee counts wasn’t considered the best messaging for broadening the appeal of apps to prospective customers. “It is tempting to conclude that occupations with high overlap will experience job loss,” they write in one of those hedges on AI’s likely impact on employment. “This would be a mistake, as our data do not include the downstream business impacts of new technology, which are very hard to predict.” On the one hand, authors do specify that AI tools are being used most often for communications tasks like language interpretation, emailing, and composing marketing materials. But on the other, they hasten to add it isn’t clear apps are also being asked to assume the complete array of tasks those workers perform on the job — or whether they’d even be capable of doing that. The researchers similarly note the different objectives employees studied had in using bots. Some of those workers asked apps to entirely handle and complete certain job tasks. But in many other cases, people queried AI assistants about the most effective ways to fulfill work duties themselves, retaining their own value to employers. Meantime, the study’s authors also seem to balance between the higher productivity objectives — and potentially decreased labor costs — that some employers hope AI will provide and employees’ contrasting fears about their job security. Those diverging focuses, the authors say, won’t generate the zero-sum results many warn of — at least not necessarily. “For example, if AI makes software developers 50 percent more productive, companies could raise their ambitions and hire more developers as they are now getting more output per developer, or hire fewer developers because they can get the same amount done with fewer of them,” they say. “Our data is only about AI usage and we have no data on the downstream impacts of that usage, so we only weigh in on the automation versus augmentation question by separately measuring the tasks that AI performs and assists.” But the study also makes it clear that the jobs least likely to be disrupted initially by increased or dominant use of AI are those involving some mixture of manual activity, use of machines, and interaction with people. That combination leaves nursing assistants, hazardous waste removers, car windshield installers, and medical equipment preparers among the professions with the lowest level of app penetration. But even there, the authors create some wiggle room for eventual employment outcomes. They note their research is based on use of AI that Microsoft developed from large language models (LLMs). More focused apps tailored to individual professions — possibly paired with robotic machines — might still leave many manual jobs vulnerable to the tech’s influence in the future. “Note that our measurement is purely about LLMs,” the authors note. “Other applications of AI could certainly affect occupations involving operating and monitoring machinery, such as truck driving.” Meaning, any historians or brokerage clerks feeling fearful about their work after reading the study might want to rethink any plans about rushing into careers as a roustabout or packaging machine operator. Because those professions, too, may come under the growing influence of specialized AI apps in the not too distant future. BY BRUCE CRUMLEY @BRUCEC_INC