2013 · Stewart Butterfield, Tiny Speck

Tiny Speck spent four years building Glitch, a quirky multiplayer game that never found a paying audience. To coordinate their distributed team they had built an internal chat tool. When the game failed, Butterfield noticed the team would not stop using the tool — they were actively grieving losing it. He shut Glitch down and shipped the tool as Slack. Within a year it was the fastest-growing SaaS product in history.

The principle: What you built to make the main project possible may be more valuable than the main project. Watch what your team refuses to live without; the future product is often there, already in production.

1903 · Wilbur Wright, Orville Wright, Samuel Langley

In 1903 Samuel Langley had $50,000 of US government funding, a steam-powered aircraft, and the prestige of the Smithsonian. His machine plunged into the Potomac. Nine days later two bicycle-shop owners with no funding and a methodical iteration discipline flew at Kitty Hawk. Langley had the resources. The Wrights had the feedback loop.

The principle: Funding lets you build the wrong thing faster. Iteration on a real prototype, even a poor one, beats a single attempt to skip straight to the working version. The advantage is not capital — it is the rate at which you can correct yourself.

1950 · Douglas MacArthur, North Korean People's Army

While North Korean forces had pushed UN troops down to a tiny perimeter around Pusan, MacArthur staged an amphibious landing 240km behind their lines at Inchon — a port written off as militarily impossible due to its 30-foot tides, narrow approach, and mud flats. Within two weeks the supply line of the entire North Korean army had been severed and Seoul recaptured. Every planner had a reason the landing wouldn't work; that consensus was precisely why the defenders weren't ready.

The principle: The position the opponent thinks is unattackable is the one they don't defend. The reasons it's 'impossible' double as the reasons surprise will be total.

1970 · Jim Lovell, Mission Control, Gene Kranz

Two days into Apollo 13's lunar mission an oxygen tank exploded, crippling the command module. Mission Control had to bring the crew home alive using only what was already aboard the spacecraft — including improvising a CO2 scrubber from a sock, a cardboard manual cover, and duct tape. The mission failed; the crew came home.

The principle: Sometimes the constraint that defines the solution is what you have right now. The right move is not 'what would solve this' but 'what would solve this using only the components in the room'. Constraint is a creative input, not just a limit.

timeless · fungi, forest trees

Under almost every forest is a network of fungal threads — mycelium — that physically connects the roots of unrelated trees. Through it, the trees trade carbon, nitrogen, water, and chemical warnings. Older trees subsidise saplings; dying trees dump their carbon into the network as a parting gift. What looks like a competitive collection of individual trees is, in part, a single distributed organism: the wood-wide web.

The principle: Above-ground individuals can be participants in an underground commons. Optimising the visible competition while ignoring the hidden cooperation tells you the wrong story about what's actually going on.