recent research Researchers from MIT and IBM provide a new perspective on the potential for AI to replace human jobs, suggesting that humans remain economically viable in many fields.
“Beyond AI exposure: Which tasks are cost-effective to automate with computer vision?” challenges common assumptions about the pace and extent of AI integration into the research workforce.
Also read: GPU security flaw puts AI data on iPhone and MacBook at risk
Economic feasibility of AI in the workplace
This study highlights an important aspect of AI adoption: economic feasibility. Despite concerns that an AI takeover is imminent, the study reveals that the cost of training and implementing AI systems for specific tasks is often prohibitive. became. This economic barrier suggests that automation is economically viable under current conditions for only some of the roles previously identified as candidates for automation. According to the researchers, at current costs, it would be economically attractive to automate tasks related to AI vision at only 23% of a worker's wage.
“Due to the high upfront cost of AI systems, we found that it is cost-effective for companies to automate only the 23% of workers’ compensation “exposed” to AI computer vision. Ta. ”
This fact has major implications for companies considering integrating AI. Most employers will have to build their own AI systems or share their sensitive information with external vendors, incurring significant costs. The current economic environment supports the replacement of significant human labor with AI in most applications, as it is nearly impossible for other employers to do so.
Beyond 'AI Exposure', we publish a new paper that considers which tasks are economically attractive to automate. We found that companies want to use AI to only automate a quarter of the tasks that can be automated. https://t.co/I7Xq6ZPEIM @MITFutureTech @MIT_CSAIL @mit_ide
— Neil Thompson (@ProfNeilT) January 22, 2024
Gradual transition to automation
Research based on the monitoring and enforcement of business ethical practices shows that the transition from humans to AI will not be abrupt. Most organizations are currently in the study phase, weighing the potential and economic feasibility of AI for their business. This slow and steady approach reflects complex underlying dynamics, implementing a blur of technological progress from day to day while maintaining some balance between short-term and long-term economic factors.
“Previous literature on ‘AI exposure’ attempts to measure the overall potential for AI to impact an area and cannot predict the pace of this automation.”
Additionally, researchers highlight the uncertainty surrounding the future of job automation. As AI capabilities continue to evolve, predicting which jobs will be replaced and when remains a complex challenge. This uncertainty highlights the need for continued research and thoughtful consideration in policy and business decision-making.
Impact on the future of work
The MIT and IBM study therefore has significant implications for the future of work and the role of AI in it. While it cannot completely dispel concerns about AI-driven automation, it does suggest a more nuanced reality. The economic feasibility of AI deployment also plays a key role in determining the pace and scope of automation. As such, human workers may not face the immediate threat of replacement that many fear.
However, researchers caution against complacency. The evolution and incorporation of AI into different parts of the economy is an ongoing process. Understanding this evolution is critical to making informed policy and business decisions and ensuring that the transition to an AI-enhanced workforce is thoughtful and beneficial.
“…making the right policy and business decisions depends on understanding how quickly AI tasks can be automated.”