The eternal complement | OpenAI
Short summary
This essay argues that superintelligence's greatest contribution won't be brilliant breakthroughs but rather handling the unglamorous institutional work needed to turn ideas into reality. The authors sketch two futures: one where AI thinks its way to progress with minimal physical resources, another where it spends most of its time managing the vast bureaucratic and logistical machinery required to execute increasingly complex experiments.
- Frontier intelligence and institutional intelligence are complements—genius designs the blueprint, but someone has to lay the stone. History shows this complementary work has grown more elaborate, not simpler, as projects scale from Galileo's telescope to the James Webb Space Telescope.
- A "civilization of depth" would see superintelligence optimize its way around resource constraints through pure reasoning, while a "civilization of width" would require ever-larger experiments and bureaucratic coordination, making execution work more valuable than ideation.
- Even if AI surpasses humans at both thinking and doing, humans may retain value through comparative advantage in frontier insights, creative diversity, and the simple fact that not all minds can be geniuses—some must be bureaucrats.
Full summary
The core argument hinges on a simple economic concept: complements—things that increase each other's value. A better telescope makes good astronomical questions more valuable, and vice versa. The piece argues that frontier intelligence (brilliant ideas) and institutional intelligence (the unglamorous work of execution) are complements. Genius designs the monument; institutions lay the stone. The authors point out that even a civilization of a billion Einsteins would need most of them mining quarries and managing accounting. Right now, AI is easing a major bottleneck by giving people the support staff to build their own ideas. But as AI gets smarter and starts generating its own research agendas, we might flip the problem—instead of starving for good ideas, we could become starved for the capacity to execute them.
The article lays out two possible futures. In a "civilization of depth," superintelligence becomes so efficient that it can make progress through pure thought, needing less physical infrastructure and bureaucratic overhead. Think of Einstein reasoning his way to relativity from minimal data. In a "civilization of width," the opposite happens: complexity grows faster than intelligence can economize around it. A superintelligence might design a century of experiments in an afternoon, then spend the actual century waiting for labs to build things and factories to produce them. History leans toward width—Galileo's telescope fit in his hands, but the James Webb required a civilization. If width wins, the real superpower of superintelligence wouldn't be genius. It would be willingness to be the bureaucrat, coordinating the vast machinery needed to turn ideas into reality.
The authors hedge on which path we'll take, but they're clear about what it means for humans. We might keep value by being relatively better at frontier thinking than institutional coordination, even if machines beat us at both. Or we could offer creative diversity—machines staying in their own self-sustaining apparatus while humans deploy a fraction of machine intelligence at their own creative behest. The unstated implication: much of whatever intelligence exists in the future, human or machine, will spend its time on boring institutional work. That's not a bug. It's the support system progress actually needs.