Artificial intelligence could impact the work tasks of roughly 80 percent of the U.S. workforce, according to new research, as the tech develops in ways that could augment or even displace some labor.
Researchers from OpenAI, OpenResearch and the University of Pennsylvania argue in a new paper, posted to arXiv, that around 80 percent of the workforce could have at least 10 percent of their work affected by the implementation of large language models, including Generative Pre-trained Transformers (GPTs), in their industry.
Nearly 20 percent of workers could see at least 50 percent of their work tasks impacted — and around 15 percent of all work tasks in the U.S. “could be completed significantly faster at the same level of quality” with access to the artificial intelligence tool.
The researchers used AI to get to their findings and looked at exposure to the tech “as a proxy for potential economic impact” and didn’t distinguish “between labor-augmenting or labor-displacing effects.”
Several occupations — including mathematicians, writers and authors, journalists, web and digital interface designers and tax preparers — as highly exposed, meaning the researchers predict GPT tech could significantly cut down work task time in the jobs.
The paper concludes that LLM technologies “could have considerable economic, social, and policy implications” if implemented in the U.S. workforce.
OpenAI developed the language processing system ChatGPT, which used advanced tech to generate human-like conversation, and its growing popularity has sparked both interest and concerns about the possible broader implications of using artificial intelligence in school, in the workforce and elsewhere.
Some have raised worries AI could be abused by students in school settings to generate essays with information pulled from the internet. Last week, fabricated AI-generated images of former President Trump being arrested circulated on the internet.