OpenAI essay argues AGI's value lies in routine work
Two OpenAI staff argue that AGI may create most of its value through routine execution rather than genius, because frontier ideas depend on the institutions and labour needed to carry them out.
OpenAI has published "The eternal complement," an essay by Hemanth Asirvatham and Elliott Mokski. It is the first in a series on "the next economy," which runs on what the company calls a new platform for independent voices exploring an AGI future. The authors note that the essay reflects their own views, not those of OpenAI or their colleagues.
The central argument comes from economics. Frontier intelligence and the capacity to execute on it are complements: more of one makes the other more valuable. The authors suggest humanity's ideas may have "outrun our execution capacity." They contrast Galileo's telescope, which "called on a few dozen hands," with the James Webb Space Telescope, a ten-billion-dollar project that drew on 300 organizations across 14 countries.
To support the point, they cite work by Nick Bloom and coauthors on research productivity. Sustaining Moore's law now takes more than 18 times as many researchers as it did in the early 1970s. Since the 1930s, effective research effort across the economy has risen 23-fold, while measured research productivity has fallen by a factor of 41.
The authors call the work of execution "institutional intelligence." They say even "a billion Einsteins" would need most of their number to "mine the quarries and manage the accounting." They argue AI is already making execution less scarce by writing code, searching literature and building prototypes. In their view, that shifts the scarce input toward taste: deciding what is worth making.
The essay contains no new model, data or product announcement. The available text is cut off before the authors finish answering the question they set out: where AI will spend most of its time.