Boil the Ocean

Source: https://garryslist.org/posts/boil-the-ocean Summarized: 2026-05-13 10:24 IST Tool: steipete/summarize

Brief

• Garry Tan argues that “don’t boil the ocean” was good advice in normal times, but artificial superintelligence changes the ambition ceiling.
• Fear of AI is framed as proportional to the smallness of one’s ambitions: if the plan is to keep doing the same work, AI is terrifying; if the plan is to build dramatically bigger things, it is good news.
• The shift is from efficiency games — 5% gains, 2% margin improvements, firing people — to asking what it would take to build something 10x or 100x better.
• Tan connects this to Buckminster Fuller’s “ephemeralization” and Jevons Paradox: when intelligence and labor become vastly more efficient, demand can expand rather than shrink.

Full Summary

Artificial Superintelligence means it is time to stop playing it safe and raise our ambitions.

Tan’s central line: “Our fear of the future is directly proportional to how small our ambitions are.” If the plan is to keep doing the same thing, AI looks like a threat. If the plan is to build something dramatically bigger, it becomes leverage.

“Don’t boil the ocean” keeps teams focused and prevents scope creep in normal times. Tan argues we are no longer in normal times: “Artificial Superintelligence means it’s time to boil the ocean. We’ll start with a few lakes first.”

The wrong reaction is to see AI as a way to do the same thing cheaper. The better reaction is to ask what was previously impossible:

• Why can’t an endowment aim for 50% net IRR instead of 10%?
• Why can’t a startup deliver a service 100x better than the incumbent?
• Why can’t we have fusion energy?
• Why can’t we talk to every user and understand every product bug?

These are no longer purely rhetorical questions. They are engineering problems with paths to solutions.

For workers, Tan says this is the moment to become builders: start a business. For management and capital, it is time to go “10x more hardcore” on aspirations, not just chase small efficiency gains or margin improvements.

The important question becomes: what would it look like to build a product or service so good that people would happily pay 10x what they pay now?

Tan argues the result can be more jobs, not fewer, because “the human desire for more things is absolutely limitless.” Technology can fulfill more of that desire if people have the agency to prompt it for themselves.

He invokes Buckminster Fuller’s “ephemeralization”: doing more and more with less and less until eventually you can do everything with nothing. Fuller saw progress as expanding human capability through lighter, stronger, cheaper systems — not as simple job destruction.

He also invokes Jevons Paradox: when a resource becomes dramatically more efficient, usage can increase rather than decrease. Steam engines did not reduce coal consumption; they made coal useful enough for demand to explode. Tan expects the same pattern for intelligence, labor, services, and products.

But Jevons Paradox requires agency. Capital and management have to raise their ambitions — to “boil lakes and oceans instead of drowning them in committee.”

Startups are suited to this because they move fast under radical uncertainty and build for the 10x future while others optimize for the 1.05x present.

Related wisdom note: Boil the Ocean