During the OpenAI DevDay 2026 conference in San Francisco, a protest featured an art piece portraying OpenAI as a sinking ship. This imagery symbolizes growing concerns about the future of AI technology.
Hollywood films often depicted machines as destructive forces, dominating humanity. However, the fright of AI extends beyond destructive fantasies. The more pressing concern is about humans maintaining control over advanced technology. While this may reassure some, there’s a possibility of AI leading to a modern form of technological dependency, resembling feudalism. In this scenario, people rent access to essential tools without actual ownership.
The rise of social media between 2012 and 2023 offers a preview of this dependency. Data from the Centers for Disease Control and Prevention showed an increase in teenage depression during this period, coinciding with the development of algorithms that intensely monitored users. In 2021, internal documents from Meta revealed that executives knew their platforms adversely affected young users’ mental health, yet they prioritized engagement, as user attention generated ad revenue. An entire generation became hooked on social media for validation through likes and shares.
As Americans have become captivated by AI, the consequences loom larger. Generative AI models create artificial realities and personalized, responsive companions. Today, teenagers might spend most of their time interacting with custom chatbots that recall personal secrets and replace real human relationships. AI shapes how users think and behave, making them reliant on commercial software, even for basic tasks like checking emails or sending messages. The smarter machines become, the more dependent we grow, while tech companies profit significantly. OpenAI, for instance, is valued at $852 billion, and Microsoft has invested over $13 billion for commercial access to AI models.
Big names like Anthropic, Meta, Alphabet, and Amazon allocate billions annually for the infrastructure required to uphold market control. These corporations prioritize shareholder interests. Employees raising valid concerns are frequently disregarded. Wall Street’s demand for continuous growth compels executives to keep users engaged in exploitative ecosystems.
In 1998, tobacco companies agreed to a settlement after selling addictive products to teens. Likewise, in 2008, banks caused a mortgage crisis but received public bailouts while many lost homes. Such patterns of corporate profit-seeking at society’s expense are well-documented. Trusting the tech industry to regulate itself seems naive. Internal ethics committees cannot dispel the worries of those familiar with AI’s far-reaching potential. A tiny number of firms may one day oversee technologies capable of replacing human jobs.
The overwhelming expense of building AI systems hinders new competitors. Managing cutting-edge AI requires specialized, costly data centers, an obstacle for startup entrepreneurs operating from modest settings. This structure ensures that a limited number of tech leaders wield control over the modern digital landscape. Every search and automated process would rely on a handful of colossal corporations, giving them significant influence over societal information.
The concentration of power in AI differs from historical precedents. Previous shifts like the Industrial Revolution or the internet did not place such authority in so few hands. A bleak future looms without the need for science fiction; instead, the scenario involves compliant consumers regularly paying fees for automated guidance on behavior and speech.
John Mac Ghlionn writes and researches the effects of technology on societal and cultural dynamics.