AI Was Supposed to Replace Workers. It’s Not Working

This video examines a growing trend where large corporations, having aggressively implemented AI to replace human workers, are now reversing these decisions. The analysis shows that AI tools often underperform in real-world, messy environments compared to controlled demo settings. Major examples include Starbucks scrapping an AI inventory system and McDonald's removing AI-powered voice ordering, both due to persistent accuracy issues that forced staff to perform redundant work. The video argues that these failures stem from AI's inability to handle complex, emotional, or edge-case situations that require human judgment, causing customer dissatisfaction and employee burnout. Furthermore, the analysis highlights that for many companies, the high costs of AI tokens and infrastructure actually exceed the savings from workforce reductions, a phenomenon exacerbated by the high failure rate of AI pilot programs. The discussion concludes that while AI can effectively automate routine tasks, it often acts as a net cost-adder rather than a saver when it forces businesses to 'clean up' its mistakes, leading to a broader shift in how companies value AI versus human labor.