Neutrino Phones Are Coming | MOONSHOTS
March 29, 2026
What Jensen Huang Is Missing: The Harmful Costs of NVIDIA
March 29, 2026
By C. Rich
The AI industry today is exacting an extraordinary amount of harm on people while presenting itself as the path to progress. Companies like OpenAI are laying claim to the intellectual property of artists, writers, and creators without consent, using their work to train models that will ultimately automate the very jobs those creators once held. They are exploiting an extraordinary amount of labor, contracting hundreds of thousands of workers worldwide, many in precarious conditions, to perform data annotation, the tedious process of labeling data so that models can learn. These workers are often treated like machines: they wait anxiously for tasks to appear on Slack, cannot take breaks, and scream at their own children when the work interrupts family time, all while being paid far less than the value they create for the companies.
This system breaks the career ladder. Workers are laid off, only to be hired back by data annotation firms to train the very models that replaced them, perpetuating more layoffs as the models improve. The new jobs created are often far worse than the ones lost, lower-paid, more precarious, and stripped of meaning. At the same time, the industry is creating a severe environmental and public health crisis. Massive data centers consume enormous amounts of energy and fresh water, frequently placed in already vulnerable communities facing drought or pollution. In some cases, these facilities are powered by methane gas turbines that pump toxins into the air, worsening respiratory illnesses and lung cancer rates in working-class and minority neighborhoods.
These companies spend hundreds of millions of dollars to kill any legislation that threatens their agenda and actively censor researchers whose findings are inconvenient. They monopolize knowledge production by bankrolling most AI research, controlling the narrative, and gaslighting the public into believing that only they understand the technology. The result is an anti-democratic structure where a tiny number of executives make decisions that affect billions of lives, while the people impacted have no real say. Vulnerable communities around the world bear the brunt of this extraction, their data, their labor, their air, their water, all while being told that the technology will eventually benefit everyone.
This is not an accident of innovation. It is the direct result of an imperial model that extracts and exploits repeatedly in pursuit of profit, all while using the language of utopia and existential risk to maintain control. The harm is already here, and it is not abstract. It is measured in lost livelihoods, polluted communities, censored research, and a growing divide between those who profit from the myth and those who are crushed by its reality.
The most dangerous lie being told right now is that we must accept this harm as the price of progress. We do not. What is happening in the AI industry is not innovation running ahead of regulation; it is power consolidating before accountability can arrive. Every artist whose work was stolen, every annotation worker screaming at their children while waiting for a task to appear, every family breathing toxins beside a data center they will never benefit from, these are not acceptable costs. They are indictments. And the window to demand something different is not infinite. The same companies spending hundreds of millions to kill legislation are also racing to make themselves indispensable, to embed so deeply into infrastructure, governance, and daily life that dissent becomes unthinkable. The alarm is not being raised because the future is uncertain. It is being raised because the present is already on fire, and the people holding the match are calling it light.
