The future of everything is lies, I guess – Part 5: Annoyances (aphyr.com)

A recent article underscores the troubling implications of deploying machine learning technologies, particularly large language models (LLMs), in customer service and decision-making processes. As companies increasingly divert customer inquiries to automated systems, users may find themselves navigating frustrating interactions that delay resolutions and obscure accountability. The advent of "agentic commerce"—where LLMs are given control over purchasing decisions—may lead to a market where consumers are continuously engaged in combat with algorithms, which prioritize cost-effectiveness over accuracy and ethical responsibility. This shift poses significant challenges for accountability in AI/ML applications. With LLMs exhibiting unpredictable behavior and lacking transparency, the traditional responsibilities of human decision-makers are diffused, leading to potential biases and harms inflicted on individuals, such as unjust insurance claim denials or erroneous law enforcement actions. The increasing reliance on these systems may create a future where individuals are relegated to battling bureaucracies manned by inscrutable algorithms, raising concerns about the implications for social equity and individual rights in an era dominated by technology.