Philosophers in AI Companies Are a Distraction
Something strange happens when you walk into a modern AI company. You pass through security, past the rows of engineers staring at code, and there, in a corner office with glass walls, sits a philosopher. This is not a joke. Google DeepMind employs Iason Gabriel, a moral philosopher and political theorist, to think about what AI should do. [1] The Guardian profiled him in 2024, asking whose moral compass should guide artificial intelligence. [2] It seemed like a reassuring sign that the people building the technology are taking ethics seriously. But if you look closer, the philosopher is less a compass and more a decoration. The infrastructure of ethical AI has become invisible because it is everywhere — every major lab has an ethics board, an AI safety team, a principle of fairness. Yet the machines keep doing what they do, and the real decisions keep being made elsewhere.
The Invisible Philosopher
The philosopher inside DeepMind is supposed to represent something noble: a commitment to reflect before we leap. Gabriel talks about a “deep mystery of what, actually, is this thing” when asked about AI consciousness. [2] He wrestles with questions of moral status and alignment. But his presence also serves a quieter function. It signals to regulators, journalists, and the public that the company is thinking hard about right and wrong. The New York Times noted in December 2024 that we can debate the ethics of AI but seem unable to change course, pointing to a crisis of democratic involvement. [3] The philosopher is a symptom of that crisis: a visible gesture toward debate that never actually alters the path. When the camera is on, the philosopher speaks. When the camera turns off, the engineers optimize for profit, for growth, for military contracts. The philosopher’s office is a stage. The infrastructure of ethical scrutiny is there, but it is hollow. It suggests that someone is steering when in fact the steering wheel is disconnected. The real decisions are made by revenue targets and national security interests, not by moral reasoning.
The Structural Cause: Incentives Over Morals
Why does this happen? The answer is not that AI companies are evil or that their philosophers are insincere. The cause is structural, embedded in the very logic that funds the technology. Hundreds of billions of dollars have poured into AI because it promises commercial returns and geopolitical advantage. Google competes with Microsoft, China with the United States, startups with every incumbent. These pressures do not wait for a philosopher to finish a paper. They accelerate development automatically. In 2010, a thought experiment called Roko’s Basilisk imagined a future superintelligent AI that would punish those who did not help create it, creating a perverse incentive to build faster. The real basilisk is not a future machine. It is today’s economic logic. The compulsion comes from competition, from the IPO clock, from military AI budgets that have no ethical sunset clause. Society has not consciously debated where AI should go. The direction was set by the architecture of venture capital and arms races. The democratic involvement that the New York Times called for remains absent. Citizens are not asked whether they want AI to help restore ecosystems or simply extract more oil. The question is already answered by the incentives. The philosopher’s public musings become a distraction from the fact that the train has left the station without a ticket for the public.
Practice vs. Promise: The Sacking and the Weapons
Now look at the detail that shows how far practice is from the promise. In 2025, Google reversed its ban on AI weaponry. [4] The company that houses a moral philosopher decided that building weapons is acceptable. Meanwhile, employees who raised internal concerns about this decision were fired. In the United States and the United Kingdom, Google retaliated against staff who questioned the ethics of the company’s actions. One such employee claims he was unfairly sacked. When The Guardian asked Iason Gabriel about these concrete ethical failings, he declined to comment. The philosopher who speaks about the deep mystery of AI consciousness had nothing to say about missiles that might be guided by the same software he helps design. The founders of DeepMind brushed off questions without providing further comment. This is not evasion. It is a strategy. The philosophical sheen applied by big tech companies is PR, not principle. It distracts attention from the everyday harms that are already happening — labor exploitation, surveillance, weapons — by pointing to lofty abstractions about the future. The real ethical failure is that the philosopher exists precisely to make these questions seem answered when they are only postponed. The infrastructure of ethical AI is a stage set. The actors have lines, but the script is written by incentives. Wisdom becomes the scarce resource, and it is not found in the corner office. It must be built by a public that refuses to let the train run without a destination.
Sources
2. The Guardian
4. Google