AI Too Expensive to Replace Many Jobs, MIT Study Finds

Only 23% of Vision Tasks Worth Automating Due to High Upfront Costs, Suggesting Gradual AI Job Displacement

New research from MIT suggests that AI is currently too expensive to replace human workers in many jobs, particularly those that require computer vision.
The study found that only 23% of the wages paid for vision tasks would be worth automating due to the large upfront costs of AI systems.
AI automation is expected to be a gradual process, providing time for policy initiatives and retraining programs to mitigate potential unemployment impacts.
Even with rapid decreases in cost, it would still take decades for computer vision tasks to become economically efficient for firms.
The researchers argue that the study's findings may also apply to generative AI or the automation of language-related tasks.