The U.S. BLS on AI & work
The BLS actually has a pretty good writeup of the reality-based changes in employment we've seen from AI, and their own estimates of likely impacts in the future. Probably a better read than a lot of the noise out there:
"AI technology continues to advance. In addition to large language models (LLMs), new iterations of image, sound, and video generation software are being deployed. This technology has generated widespread public interest, including its potential implications for the future labor market.
BLS projections methods include some key assumptions which frame the expected employment impact of AI. One is the assumption of a full-employment economy in the projected year.6 This is assumed in order for the projections to ignore cyclical economic fluctuation and focus on long-term structural change in the distribution of employment among occupations and industries. This assumption is part of the overall labor force and macroeconomic components of the projections model, which projects total employment growth of 3.1 percent from 2024 to 2034.
Second, the projections assume that the pace of structural change in the economy due to technology will follow its historical pattern. This has traditionally been gradual, with job displacement due to technological change typically taking longer than technologists expect.7 While it is possible that future technologies may affect the labor market faster than has been the case for past technologies, BLS projections methods are not designed to account for rapid change. Therefore, BLS assumes that the overall pace of technological change will be consistent with past experience.
However, AI is already being deployed in a variety of work settings, with likely productivity-enhancing effects on a range of occupations. Several occupations are projected to experience employment impacts because of the increased adoption of AI, including generative AI (which can create new content rather than just use existing data to do analysis). The following examples should not be considered exhaustive or definitive. Rather, they are examples in which a reasonable expectation of an AI-driven impact currently exists. Because the capabilities of AI systems continue to evolve, these assessments can change. Generative AI also carries an array of risks, such as those related to hallucinations, training data bias, and intellectual property infringement, among others.8 Thus, there is considerable uncertainty around the extent of generative AI adoption across the economy and its impact on employment. BLS releases new projections data annually and, as more information emerges that helps clarify the extent of the technology’s impact, it will be incorporated into future sets of projections data.
Arts, design, media, and communication occupations are expected to be particularly susceptible to productivity effects from generative AI. Graphic designers, for example, can leverage generative AI tools in the initial step of the design process, which allows these workers to explore and iterate ideas faster, enabling them to create first drafts much more quickly and to shift their focus to later stages of the design process. Generative AI can also be used to automate repetitive tasks and speed up certain processes, such as prototyping and design verification.9 These productivity gains are expected to limit demand for this occupation, which is projected for slower than average employment growth for 2024–34.10 Meanwhile, interpreters and translators have become more productive as AI translation capabilities continue to improve, reducing their employment demand.11 Other examples of occupations in this group whose tasks have a high potential to be automated or streamlined by generative AI systems include special effects artists and animators (as video generation software improves in speed and quality), technical writers (as documentation is produced more efficiently), and broadcast announcers and radio DJs (as adoption of AI-generated voice technology increases). Therefore, the technology is expected to limit the employment demand of these workers between 2024 and 2034.
Several healthcare-related occupations are also expected to be affected by AI. For example, medical secretaries and administrative assistants are expected to see productivity-enhancing effects on their billing and claims management tasks.12 AI-based tools that make the medical-coding process more efficient are also expected to moderate demand for medical records specialists. Among healthcare practitioners, radiologists are incorporating AI into traditional computer-aided diagnosis (CAD) systems, which increases efficiency.13 However, these productivity effects are counteracted by the strong underlying demand for healthcare, meaning employment growth is still projected for these occupations. Employment of medical transcriptionists, however, is projected to decline 4.9 percent from 2024 to 2034 because AI technology can recognize speech and transcribe audio, reducing the need for these workers.14
AI technologies are also expected to affect a number of sales occupations. For instance, AI systems can quickly summarize sales calls, analyze their content, and recommend next steps based on the information exchanged. Moreover, AI solutions can assist sales representatives with lead prospecting and scoring as well as with researching prospects and the competition.15 Generative AI tools can also be employed to draft emails, create presentations, and develop sales proposals—tasks that salespeople routinely engage in.16 By reducing the time and effort spent handling manual and repetitive tasks, AI allows sales representatives to focus more on engaging customers and closing deals and reduces their employment demand. Relevant occupations that are expected to see productivity improvements and decreased demand for their labor as a result of AI integration include sales engineers, insurance sales agents, and other types of sales representatives.
Automation technology has long been a factor impacting the job outlook of many office and administrative support occupations, with the productivity gains associated with the adoption of digital tools, such as automated phone systems and virtual assistants, constraining demand for these workers. As the integration of existing and new AI technologies into workflows expands, various types of office and administrative support workers are expected to see additional efficiency gains. As a result, demand is expected to be limited for occupations such as billing and posting clerks; procurement clerks; credit authorizers, checkers, and clerks; customer service representatives; and nonmedical secretaries and administrative assistants. Employment of these occupations is projected to decline or show little change over the 2024–34 decade.
A variety of other jobs are also expected to be impacted by AI. For example, AI-powered tools designed specifically for the legal services industry can help paralegals and legal assistants review contracts, streamline the discovery process, and conduct research.17 The efficiency gains resulting from these and other AI solutions are likely to constrain job growth, with the paralegals and legal assistants occupation projected to see little or no employment change over 2024–34. Likewise, digital tutors can use AI and LLM technology to impart information and answer student questions, which is expected to slow employment growth for tutors.18 Claims adjusters, examiners, and investigators can use AI tools to assess photographs of property damage, which improves efficiency in generating payout estimates.19 Consequently, employment of these workers is projected to decline 5.1 percent from 2024–34. (See table 2.)
Although several occupations, including but not limited to those discussed above, are expected to be affected by the labor-saving effects of AI adoption, others are expected to experience a positive impact to their employment outlook. This is the case for some computer occupations, whose skills will be increasingly needed to meet the growing demand for AI-based systems, which are discussed in more detail in the IT section."
snowghoul
origamiaround
yousharknotpass
kianamaiart
blakekathryn
titsay
areudirty
