Why Some AI Projects Are Worth 100 Times More Than Others

An essay written to a colleague at Palisade explains why the author feels unsatisfied with projects that others on the team are more enthusiastic about. It uses the example of spending six months making 100,000 bricks toward the 70 million needed for a wall. If those bricks help build kilns that make many more bricks—putting you upstream of an exponential process—the project was leveraged and potentially hugely important. If instead a brick-making school scales up to produce thousands of brick-making teams, you are downstream of an exponential; your contribution was noble but likely misallocated, because the school either overshoots the target or falls far short. And if neither happens, your bricks didn’t amount to anything important. The real strategic question is whether your work feeds a compounding process.