Member-only story
A 1981 Wargames Strategy Applied To Investing
Portfolio construction gleaned from an A.I. guided victory
On July 4, 1981, a 29-year-old Stanford University computer scientist named Doug Lenat entered the Traveller Trillion Credit Squadron tournament in San Mateo, California. Contestants were allocated an imaginary budget of a trillion dollars to build warships that partook in battle via knockout rounds to determine the winner. Most of the contestants’ fleets resembled those used in real-world conventional warfare, consisting of variously sized warships capable of defending themselves.
Lenat adopted a different strategy.
He fed the tournament rules to an Artificial Intelligence program he developed. After some tweaking of the parameters, it guided him to deploy an enormous flotilla of tiny ships with ample firepower but entirely devoid of defensive ability. Lenat’s ships were sitting ducks opponents could easily sink — his opponents just couldn’t sink all of them before being overwhelmed by his flotilla’s cumulative firepower. Lenat easily won the tournament.
As an investor, I find stories of conventional wisdom turned on its head intriguing as they often contain lessons applicable to my investment philosophy. Lenat’s 1981 AI-guided gamesmanship is an example of utilizing pattern recognition to…