Creative Optimizers
Most people who have the intention to change the world fail to do so. They overestimate their abilities, underestimate the difficulty of the challenge, or both.
One group of people, Creative Optimizers, are remarkable because of the unique way they face problems. And while their efforts are often resisted, everyone looking to change the world in big ways needs to study these people.
Examples from Games
The most obvious examples of Creative Optimization are in games. Here are three striking examples:
AI in tabletop gaming — In 1981, using the networked computers at Xerox Parc, Doug Lenat trained an AI to play a popular tabletop wargame that he himself had never played. His AI designed fleets were unlike anything the human competitors created, with large numbers of tiny, highly armored ships and many other innovations. Lenat won the national tournament two years running with numerous opponents conceding without a fight.
An artists rendering of “Traveller: Trillion Credit Squadron”, the game Lenat dominated in and which is so complex that even basic gameplay requires referencing large data tables.
Physics in golf — Even before turning pro in 2016, Bryson Dechambeau was rethinking golf using physics. He has since adopted a highly mechanical, consistent swing, gained >20lbs of muscle, and made other changes to create an ugly and highly effective new style of play. He can now drive the ball further and more accurately than almost any other player; in 2021 he was the 4th ranked professional golfer in the world.
Reading Homer Kelley’s 1982 book “The Golfing Machine” at age 15 was a pivotal influence on DeChambeau’s foundational approach to the game.
Math in blackjack – In 1961, MIT math professor Edward Thorp developed a method to systematically beat casinos in blackjack; he called it “card counting.” Thorp used this method to win $210,000 (in 2023 dollars) in a single weekend; card counting is now world-famous and has been the subject of numerous Hollywood movies.
Edward O. Thorp’s subsequent book “Beat the Dealer” has sold over 1 million copies since it was published in 1966.
Even though the domains are different and the methods are different, these cases have a common thread:
— In each case, the person violates the spirit of the game without breaking the rules of the game. This allows them to search through novel solutions until the person finds a radically new, highly effective approach. —
The above examples are from traditional “games,” and these simple examples make it easy to understand the concept. Creative optimization, however, occurs in many domains, including in places where the outcomes are very important.
Examples from Life
Creative optimization outside of games can be harder to spot for two reasons. First, the “game,” both the rules and its spirit, is less clearly defined. Second, successful new ways of doing things are often quickly adopted and absorbed into “the way things have always been done.” As a result, you may have to pay close attention to notice the process of creative optimization is at work.
Here are some real life examples, where the idea or people violate the spirit of some game with a surprising and effective new strategy:
Smallpox eradication — In 1966, D. A. Henderson and his team were tasked with eradicating smallpox, one of the deadliest diseases in human history. The idea of eradication was opposed or considered impossible, yet after ten years the World Health Organization’s Smallpox Eradication Campaign, led by Henderson, completely eliminated naturally occurring instances of smallpox.
Smallpox eradication involved many further examples of creative optimization; for more, see Smallpox: Death of a Disease.
Free samples — In the 1850s, Benjamin T. Babbitt started giving out free samples of his soap. This idea was shocking, since the point of commerce and business was to make money, and yet Babbitt was giving his product away for free. Today, the idea of free samples is commonplace and considered obvious. Babbitt saw tremendous success with his Lower Manhattan factory covering 300,000 sq. feet and producing 100,000lbs of soap per day.
Babbitt’s was also one of the first companies to create colorful advertisements.
Gerrymandering — In 1812, Democrat-Republican governor (and later vice president) Elbridge Gerry helped carve Massachusetts’ electoral districts in a highly unnatural, yet perfectly legal way in order to swing elections in his party’s favor. The strategy succeeded in letting the Democratic-Republicans maintain control of the state senate.
The original Gerry-mander was named for Elbridge Gerry and the salamander-like shape of the redistricting.
Even though eradication, free samples, and gerrymandering seem like obvious ideas now, at the time they were extremely novel. In all cases, some supposed ‘spirit of the game’ was violated, but rules of reality, capitalism, and politics were obeyed. The fact that the solutions were so surprising and so effective may even indicate the ideas could have been developed much earlier — whether they should have or not, as in the case of gerrymandering. Arguably never would've been the best time to invent gerrymandering.
Costs and Impacts
Creative optimization can yield highly effective new approaches. But it also can be costly in several different ways. The extensive search for new methods can take a lot of time — Lenat took 6 years to design and then train his AI, Thorp took 2-3 years to develop card counting, and Dechambeau spent more than 7 years on his style before turning pro in 2016.
More importantly, creative optimizers frequently face substantial opposition.
De facto banning — Doug Lenat was laughed at and then, when he started winning, had several players appeal to the judges to get him disqualified. After he won a second year in a row, the tournament organizers told Lenat that they wouldn’t hold the competition next year if he showed up.
Ridicule — Bryson DeChambeau is widely considered the most hated golf player in recent history, with rival golfers describing his approach as “embarrassing” or “not a skill.”
Backlash — Elbridge Gerry received sufficient (justified) criticism for his role in redistricting Massachusetts that he lost his seat as governor in the same year the Democratic-Republicans maintained control of the state senate. This was despite his claimed position that partisan redistricting was "highly disagreeable”.
Poisoning and sabotage — Edward O. Thorp was poisoned twice and his car was sabotaged, which he discovered while driving 65 mph. He and the four other passengers narrowly escaped with their lives.
Incredulity — The smallpox eradication campaign was thought to be impossible by most experts including the director of the WHO from 1953-1973 who: “...opposed a global smallpox eradication campaign that he believed would fail and damage WHO’s credibility.”
Whether the costs and challenges are worth it depend on the nature of the effect and the degree of its adoption. Some instances of creative optimization are grudgingly accepted, like the filibuster (first known instance in 60 BC) or gerrymandering (first known instance in 1812). Others are adopted seamlessly, like free samples or disease eradication, once they are proven to be possible or to work.
Isn’t this just innovation?
Creative optimization is innovation, but a very specific kind. Most innovation is incremental; a good example is the smartphone. The first touchscreen phone was the IBM Simon (released in 1992), which built off the first cell phone, which was the Motorola DynaTAC 8000X (released in 1983), itself made possible by the touchscreen in 1965.
While invention requires many innovations, it doesn’t necessarily require Creative Optimization.
Creative optimization is much more revolutionary. It contains an element of surprise and instant disruption; it expands the notion of what is possible. While the majority of value produced is through incremental innovation, there are times when a Creative Optimizer shows up and blows things wide open.
Should you always be Creative Optimizing?
Creative optimization is needed in some places but not all. Do we need creative optimization in the case of food? Probably not. (Though see Soylent.) In some cases, things already work. In other cases, only incremental changes are needed. In still other cases, the extent of the backlash is prohibitive. Judgment is needed to determine when creative optimization is the right choice, and when it isn’t.
Conclusion
Creative optimization is powerful, costly, and sometimes makes the world much better. To learn a bit about how to be a creative optimizer, check back later for the next essay in the series.