Every new technological advancement throughout history has triggered concerns about job security, and the rise of artificial intelligence (AI) is no exception. With cutting-edge technologies like OpenAI’s ChatGPT and Google’s Bard, capable of performing diverse tasks with a single text prompt, fears about job displacement have intensified.
Even the creators of these technologies, including OpenAI’s Sam Altman, have acknowledged the potential transformation of the job market as generative AI technology continues to advance.
Previous estimates by Goldman Sachs indicated that AI could automate around 25% of the global workforce, with McKinsey projecting nearly 50% of all work to be AI-driven by 2025. A joint analysis by the University of Pennsylvania, NYU, and Princeton suggested that ChatGPT alone could impact almost 80% of the workforce.
A recent 45-page analysis by the Massachusetts Institute of Technology (MIT) delves into not only the possibility of job replacement but also the economic feasibility of automating specific tasks. The study reveals that only 1.6% of worker wages in the US economy can be automated, and of those tasks, only 23% (0.4% of total tasks) are economically viable for automation.
Addressing the potential of AI technology, the study states, “”Machines will steal our jobs” is a sentiment frequently expressed during times of rapid technological change. Such anxiety has re-emerged with the creation of large language models (e.g., ChatGPT, Bard, GPT-4) that exhibit considerable proficiency in tasks previously exclusive to human beings.”
The study also highlights that, as of now, computer vision only has an economic advantage in 23% of vision tasks at the firm level. Barriers to AI-as-a-service deployments exist, and a significant reduction in the cost of computer vision is necessary for it to replace human labor. Even with a 50% annual cost decrease, the study predicts that it will take until 2026 before half of vision tasks have a machine economic advantage, and by 2042, some tasks will still favor human labor.
Regarding the potential for AI job loss, the MIT study suggests, “Our findings suggest that AI job displacement will be substantial but also gradual – and therefore there is room for policy and retraining to mitigate unemployment impacts.”
