For decades, movies like terminatorby James Cameron, we were accustomed to thinking about armies of robots since a dystopian perspectiveif you will, as an exaggeration typical of science fiction, a narrative resource to talk about fear of the future. The problem is that, little by littlethat future has stopped seeming so distant, and some of the ideas that previously only fit in the cinema are beginning to appear in the real world with a disturbing naturalness.
From the worker robot to the soldier. Most of the humanoid robot startups that have emerged in recent years sell a reassuring promise– Machines designed to work in factories, warehouses, hospitals or even homes, alleviating labor shortages and increasing productivity.
Foundationa young Silicon Valley company, shares that ambition, but takes it to much more uncomfortable terrain: his Phantom robot It is not only designed for industrial work, but also for armed combat, with the United States Army as an explicit client. Its founder, Sankaet Pathak, does not hide the intention nor the schedule: manufacture 50,000 humanoids before the end of 2027 and turn them into an operational tool for both the civilian economy and the battlefield.
Impossible calendar. They counted in Forbes that Foundation boasts an unusual development speed even by industry standards. In just 18 months since its founding, Phantom was already making real production tasks in facilities of undisclosed industrial partners, a pace comparable to that of the most advanced players in the market.
This acceleration is explained by two key acquisitions in artificial intelligence and new generation actuators, but also by a recruited team directly from companies like Tesla, Boston Dynamics, SpaceX or 1X. The scaling plan is as ambitious as it is risky: 40 robots this year, 10,000 next year and 40,000 in 2027. Pathak admits which is an extreme goal, but insists that there is a “non-zero probability” of achieving it, relying on a philosophy inherited from Tesla: do not try to automate everything too quickly.
Foundation
The economic model. The commercial bet by Foundation It is not about selling robots, but for renting them. The company isn’t looking for dozens of small customers, but rather a few gigantic contracts capable of generating hundreds of millions in recurring revenue. If the plan is fulfilled, 50,000 rented robots between 2026 and 2027 could translate into about 5 billion dollars annuallywith an approximate price of $100,000 per robot per year.
At first glance it seems expensive compared to an average human salary, but the argument is purely industrial: A humanoid can work almost 24/7 and replace between three and five people. Even discounting maintenance, human supervision and downtime, the potential savings per unit could be around $90,000 annually. All of this, of course, under a crucial condition that no one has yet demonstrated: that the robot is really as fast, reliable and versatile as a human worker.
Technology that does not exist. Phantom boasts of advanced “muscles”, efficient and reversible actuators that allow it to operate for several shifts without overheating and coexist with people with a reasonable level of safety.
Still, there is an uncomfortable reality in the sector: no manufacturer has yet achieved a humanoid that is fully equivalent to human performance in complex environments. Therefore, the money intelligent It discounts delays, reduces expectations, and assumes that it will take additional years for hardware and software to reach true maturity. The recent history of robotics is full of promises ahead of their time.
An armed robot. It is in the military sphere where Foundation definitively breaks with the comfortable narrative. Pathak defend that an armed humanoid can be “the first body in” in high-risk situations, because a docile robot does not force the enemy to reveal itself. PhantomAccording to his vision, it must be lethal.
The range of uses it’s wide: carry ammunition, perform dangerous tasks, explore buildings, cross ridges or enter caves where no officer would want to send a soldier. In fact, it is not pure science fiction: terrestrial robots have already been seen with similar functions in the Ukrainian war, although not humanoid in shape.
More precise (or easier) warfare. Foundation argues that these robots could make war more precise, not more brutal. Instead of bombing or heavy weapons, a terrestrial humanoid could evaluate situations directly. The operating model would resemble that of current drones: the robot would move and navigate autonomously, but the lethal decision would remain in human hands, remote and safe.
If that scheme works, armed humanoids could alter the logic of deterrence, substituting human deployments for robotic force demonstrations scalable. Pathak even arrives to affirm that an army with tens of thousands of visible robots could prevent wars before they start.
The ethical dilemma. There is no doubt, the other side of the argument is just as disturbing. If sending robots reduces the political and human cost of war, it can also make it more likely. History shows that when the threshold for sacrifice is lowered, resort to force becomes more tempting.
The ethics of armed humanoid robots become like this more complex than everespecially in a world where China, Russia and the United States are already developing lethal autonomous systems, even if they do not take human form. In reality, automated warfare is not new: Nazi V-2 missiles They already incorporated a primitive form of autonomy during the Second World War. What changes now is the degree of sophisticationthe distributed decision-making capacity and the physical proximity of the robot to the human combatant.
Image | Foundation
GIPHY App Key not set. Please check settings