The study determines how “applicable” AI is to certain job roles, which in theory could be used to enhance productivity therein, but in practice, will most likely be used to justify reducing team sizes \u2014 outsourcing human work to AI models. Microsoft itself is thought to have laid off over 15,000 people this year due to the “application” of AI<\/a>, and prioritizing its growth over other parts of the business. So far, there’s no evidence that AI has led to job creation.<\/p>\n You may like<\/p>\n Microsoft is at the absolute forefront of artificial intelligence, with its Azure data centers contributing processing power to models like ChatGPT<\/a>, DALL-E<\/a>, and Grok<\/a>. Its Copilot<\/a> chatbot is baked into Windows, and its business-grade solutions in Microsoft 365<\/a> aim to help users save time by offloading more menial tasks.<\/p>\n Increasingly, though, it seems that both current and future AI models will be able to not only offload those menial tasks but also replace entire professions. AI at scale could lead to a variety of unknown consequences that governments and society at large seem ill-prepared for \u2014 but perhaps there are opportunities around the corner, too?<\/p>\n In any case, here’s the list of 40 jobs most likely impacted by AI, and the 40 least likely, according to Microsoft Research.<\/p>\n Swipe to scroll horizontallyTop 40 jobs most at risk from AI (via Microsoft Research)<\/p>\n #<\/p>\n Job Title <\/p>\n Score<\/p>\n Employment<\/p>\n 1<\/p>\n Interpreters and Translators<\/p>\n 0.49<\/p>\n 51,560<\/p>\n 2<\/p>\n Historians<\/p>\n 0.48<\/p>\n 3,040<\/p>\n 3<\/p>\n Passenger Attendants<\/p>\n 0.47<\/p>\n 20,190<\/p>\n 4<\/p>\n Sales Representatives of Services<\/p>\n 0.46<\/p>\n 1,142,020<\/p>\n 5<\/p>\n Writers and Authors<\/p>\n 0.45<\/p>\n 49,450<\/p>\n 6<\/p>\n Customer Service Representatives<\/p>\n 0.44<\/p>\n 2,858,710<\/p>\n 7<\/p>\n CNC Tool Programmers<\/p>\n 0.44<\/p>\n 28,030<\/p>\n 8<\/p>\n Telephone Operators<\/p>\n 0.42<\/p>\n 4,600<\/p>\n 9<\/p>\n Ticket Agents and Travel Clerks<\/p>\n 0.41<\/p>\n 119,270<\/p>\n 10<\/p>\n Broadcast Announcers and Radio DJs<\/p>\n 0.41<\/p>\n 25,070<\/p>\n 11<\/p>\n Brokerage Clerks<\/p>\n 0.41<\/p>\n 48,060<\/p>\n 12<\/p>\n Farm and Home Management Educators<\/p>\n 0.41<\/p>\n 8,110<\/p>\n 13<\/p>\n Telemarketers<\/p>\n 0.40<\/p>\n 81,580<\/p>\n 14<\/p>\n Concierges<\/p>\n 0.40<\/p>\n 41,020<\/p>\n 15<\/p>\n Political Scientists<\/p>\n 0.39<\/p>\n 5,580<\/p>\n 16<\/p>\n News Analysts, Reporters, Journalists<\/p>\n 0.39<\/p>\n 45,020<\/p>\n 17<\/p>\n Mathematicians<\/p>\n 0.39<\/p>\n 2,220<\/p>\n 18<\/p>\n Technical Writers<\/p>\n 0.38<\/p>\n 47,970<\/p>\n 19<\/p>\n Proofreaders and Copy Markers<\/p>\n 0.38<\/p>\n 5,490<\/p>\n 20<\/p>\n Hosts and Hostesses<\/p>\n 0.37<\/p>\n 425,020<\/p>\n 21<\/p>\n Editors<\/p>\n 0.37<\/p>\n 95,700<\/p>\n 22<\/p>\n Business Teachers, Postsecondary<\/p>\n 0.37<\/p>\n 82,980<\/p>\n 23<\/p>\n Public Relations Specialists<\/p>\n 0.36<\/p>\n 275,550<\/p>\n 24<\/p>\n Demonstrators and Product Promoters<\/p>\n 0.36<\/p>\n 50,790<\/p>\n 25<\/p>\n Advertising Sales Agents<\/p>\n 0.36<\/p>\n 108,100<\/p>\n 26<\/p>\n New Accounts Clerks<\/p>\n 0.36<\/p>\n 41,180<\/p>\n 27<\/p>\n Statistical Assistants<\/p>\n 0.36<\/p>\n 7,200<\/p>\n 28<\/p>\n Counter and Rental Clerks<\/p>\n 0.36<\/p>\n 390,300<\/p>\n 29<\/p>\n Data Scientists<\/p>\n 0.36<\/p>\n 192,710<\/p>\n 30<\/p>\n Personal Financial Advisors<\/p>\n 0.35<\/p>\n 272,190<\/p>\n 31<\/p>\n Archivists<\/p>\n 0.35<\/p>\n 7,150<\/p>\n 32<\/p>\n Economics Teachers, Postsecondary<\/p>\n 0.35<\/p>\n 12,210<\/p>\n 33<\/p>\n Web Developers<\/p>\n 0.35<\/p>\n 85,350<\/p>\n 34<\/p>\n Management Analysts<\/p>\n 0.35<\/p>\n 838,140<\/p>\n 35<\/p>\n Geographers<\/p>\n 0.35<\/p>\n 1,460<\/p>\n 36<\/p>\n Models<\/p>\n 0.35<\/p>\n 3,090<\/p>\n 37<\/p>\n Market Research Analysts<\/p>\n 0.35<\/p>\n 846,370<\/p>\n 38<\/p>\n Public Safety Telecommunicators<\/p>\n 0.35<\/p>\n 97,820<\/p>\n 39<\/p>\n Switchboard Operators<\/p>\n 0.35<\/p>\n 43,830<\/p>\n 40<\/p>\n Library Science Teachers, Postsecondary<\/p>\n 0.34<\/p>\n 4,220<\/p>\n A lot of the jobs in the top 40 “most at risk” revolve entirely around what LLMs like ChatGPT<\/a> seem to be able to do best. It’s all largely impersonal, easily digitizable work that can be reproduced easily by AI. I ranked the table by the score, from most at risk to least at risk, according to the research paper.<\/p>\n I was surprised to see things like models on the list, but it makes sense when you can easily ask ChatGPT or Copilot to generate an image of a model doing anything you want with a simple command prompt. It seems Microsoft Research thinks AI will be able to even reproduce my awesome Xbox exclusives in the future too … damn.<\/p>\n Perhaps I should consider a career in phlebotomy, or become a masseuse?<\/p>\n Swipe to scroll horizontallyTop 40 jobs least at risk from AI (via Microsoft Research)<\/p>\n #<\/p>\n Job Title<\/p>\n Score<\/p>\n Employment<\/p>\n 1<\/p>\n Dredge Operators<\/p>\n 0.00<\/p>\n 340<\/p>\n 2<\/p>\n Bridge and Lock Tenders<\/p>\n 0.00<\/p>\n 3,460<\/p>\n 3<\/p>\n Water Treatment Plant and System Op.<\/p>\n 0.00<\/p>\n 120,710<\/p>\n 4<\/p>\n Foundry Mold and Coremakers<\/p>\n 0.00<\/p>\n 2,700<\/p>\n 5<\/p>\n Rail-Track Laying and Maintenance Equip. Op.<\/p>\n 0.00<\/p>\n 4,010<\/p>\n 6<\/p>\n Pile Driver Operators<\/p>\n 0.00<\/p>\n 3,010<\/p>\n 7<\/p>\n Floor Sanders and Finishers<\/p>\n 0.00<\/p>\n 4,510<\/p>\n 8<\/p>\n Orderlies<\/p>\n 0.01<\/p>\n 48,710<\/p>\n 9<\/p>\n Motorboat Operators<\/p>\n 0.01<\/p>\n 2,710<\/p>\n 10<\/p>\n Logging Equipment Operators<\/p>\n 0.01<\/p>\n 23,720<\/p>\n 11<\/p>\n Paving, Surfacing, and Tamping Equipment Op.<\/p>\n 0.01<\/p>\n 43,080<\/p>\n 12<\/p>\n Maids and Housekeeping Cleaners<\/p>\n 0.01<\/p>\n 836,230<\/p>\n 13<\/p>\n Roustabouts, Oil and Gas<\/p>\n 0.01<\/p>\n 43,880<\/p>\n 14<\/p>\n Roofers<\/p>\n 0.01<\/p>\n 135,140<\/p>\n 15<\/p>\n Gas Compressor and Gas Pumping Station Op.<\/p>\n 0.01<\/p>\n 4,400<\/p>\n 16<\/p>\n Helpers\u2013Roofers<\/p>\n 0.01<\/p>\n 4,540<\/p>\n 17<\/p>\n Tire Builders<\/p>\n 0.01<\/p>\n 20,660<\/p>\n 18<\/p>\n Surgical Assistants<\/p>\n 0.01<\/p>\n 18,780<\/p>\n 19<\/p>\n Massage Therapists<\/p>\n 0.01<\/p>\n 92,650<\/p>\n 20<\/p>\n Ophthalmic Medical Technicians<\/p>\n 0.01<\/p>\n 73,390<\/p>\n 21<\/p>\n Industrial Truck and Tractor Operators<\/p>\n 0.01<\/p>\n 778,920<\/p>\n 22<\/p>\n Supervisors of Firefighters<\/p>\n 0.01<\/p>\n 81,120<\/p>\n 23<\/p>\n Cement Masons and Concrete Finishers<\/p>\n 0.01<\/p>\n 203,560<\/p>\n 24<\/p>\n Dishwashers<\/p>\n 0.02<\/p>\n 463,940<\/p>\n 25<\/p>\n Machine Feeders and Offbearers<\/p>\n 0.02<\/p>\n 44,500<\/p>\n 26<\/p>\n Packaging and Filling Machine Op.<\/p>\n 0.02<\/p>\n 371,600<\/p>\n 27<\/p>\n Medical Equipment Preparers<\/p>\n 0.02<\/p>\n 66,790<\/p>\n 28<\/p>\n Highway Maintenance Workers<\/p>\n 0.02<\/p>\n 150,860<\/p>\n 29<\/p>\n Helpers\u2013Production Workers<\/p>\n 0.02<\/p>\n 181,810<\/p>\n 30<\/p>\n Prosthodontists<\/p>\n 0.02<\/p>\n 570<\/p>\n 31<\/p>\n Tire Repairers and Changers<\/p>\n 0.02<\/p>\n 101,520<\/p>\n 32<\/p>\n Ship Engineers<\/p>\n 0.03<\/p>\n 8,860<\/p>\n 33<\/p>\n Automotive Glass Installers and Repairers<\/p>\n 0.33<\/p>\n 16,890<\/p>\n 34<\/p>\n Oral and Maxillofacial Surgeons<\/p>\n 0.33<\/p>\n 4,160<\/p>\n 35<\/p>\n Plant and System Operators, All Other<\/p>\n 0.33<\/p>\n 15,370<\/p>\n 36<\/p>\n Embalmers<\/p>\n 0.33<\/p>\n 3,380<\/p>\n 37<\/p>\n Helpers\u2013Painters, Plasterers, …<\/p>\n 0.33<\/p>\n 7,700<\/p>\n 38<\/p>\n Hazardous Materials Removal Workers<\/p>\n 0.33<\/p>\n 49,960<\/p>\n 39<\/p>\n Nursing Assistants<\/p>\n 0.33<\/p>\n 1,351,760<\/p>\n 40<\/p>\n Phlebotomists<\/p>\n 0.29<\/p>\n 137,080<\/p>\n I ranked these jobs via the score given by Microsoft Research, with the lower the score indicating the least at risk.<\/p>\n The jobs least at risk are ones that seem to require a physical human touch, literally in some cases. Jobs like massage therapists, builders, roofers, engineers, and surgeons seem safe from AI today. But it’s certainly true that robots are becoming increasingly sophisticated and tactile as well \u2014 you’d have to think that it’s only a matter of time before humanoid robots, designed by AI itself, could supplant a lot of these roles as well.<\/p>\n It’s all getting a bit Matrix-y, isn’t it?<\/p>\n The AI upheaval will have massive consequences, for better or worse<\/p>\n Microsoft Copilot isn’t as popular as other AI chatbots, but it’s decent enough. (Image credit: Getty Images | Cheng Xin)<\/p>\n Artificial Intelligence is going to have massive implications<\/a> for the world, much like the industrial revolution did back in the 1800s. I often see investors and technologists hand-wave and apply analogies like “cars destroyed the horse and cart industry,” but increasingly, we’re seeing governments waking up to the absolute carnage this technology could wreak.<\/p>\n With our entire economic system revolving around work, and increasingly white collar work, vast unemployment, the likes of which has never been seen before, could trigger all sorts of social turmoil. Governments certainly seem ill-prepared for the AI worst-case scenario situations, even before you consider artificial general intelligence and self-replicating improvement models<\/a> that are on the horizon.<\/p>\n AI could turbocharge inequality and wealth disparity, with societies voting for populist candidates who promise a world of change while delivering more of the same. There’s only so much bullshit society can take, and AI could be the spark to tip it over the edge.<\/p>\n But what about the benefits? AI could speed up disease and cancer solutions<\/a>, able to experiment, comprehend, and parse data far faster than huge teams of scientists could manually. Perhaps AI itself comes up with economic solutions to solve the looming unemployment crisis, with some form of distributed universal basic income<\/a>, touted by some, as freeing humanity from the toils of work in general.<\/p>\n![\"CHONGQING,](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2025\/07\/JwAT2VC7ttp6PH4CqMx7fH.jpg\")
<\/p>\n