300 - Hidden Markets

Recently I've been reading three books that overlap on a particular topic, Blue Ocean Strategy Extended Edition by W. Chan Kim and Renee Mauborgne, The Infinite Game by Simon Sinek, and The Prosperity Paradox by Clayton M. Christensen. They all point to major opportunities, those who tapped into them, and those who missed them, with the latter often resulting in their bankruptcy. They each do this from their own angles, painting a more complete picture of the dynamics in play.

For example, Apple (under Steve Jobs) was faced with the dawn of the graphical user interface (GUI), after having already invested millions in a different path. They overcame the Sunk Cost Fallacy and made this pivot as part of their long-term vision, with many similar examples of that long-term thinking greatly surpassing short-term losses.

Meanwhile, Kodac and Nokia were both negative examples of failures to pivot, making mistakes that most VCs have repeated in the past 2 years. The consequences for these mistakes are every bit as predictable as a short fall off of a tall cliff, even if the specific mid-air flailing varies.

Take OpenAI and its cohort for example. They cling to a dead-end/niche technology out of Sunk Cost, among other reasons, even while they have to burn 3 to 5 orders of magnitude more compute just to compete on toy problems. With each new model, the waste continues to explode, with only marginal and usually superficial "gains", which are overshadowed by increasingly obvious benchmark fraud. They know the "Blue Oceans" that people want, they hype these up constantly and claim that they'll deliver them, even though that is very literally impossible with their technology.

They see the Blue Ocean, yet they opt to play the Finite Game (of fraud). Meanwhile, the same paradox repeats itself that was seen in the first mobile network established in Africa, where avenues of investment fail to recognize the value of "unconsumers", the people who want a thing, but have no current option that can deliver it. Trivial AI technology like LLMs, RL, and "agents" can't deliver >90% of the use cases and capacities that are hyped up, meaning that the world's largest pool of "unconsumers" in human history is silently waiting for someone to deliver, and nobody is investing in the technology that can, at least not yet.

Cognitive Biases are useful until they are wrong, but when they're wrong they are also "systematically wrong", meaning that when that happens a "dart-throwing-chimpanzee" (Prof. Tetlock's way of saying random chance) could reliably outperform the biased humans. Most of the corporate and investment domains today are systematically wrong when it comes to "unconsumers" and the "Blue Ocean Markets" they inhabit, as that population defies most common methods of measurement.

Innovation is taking calculated risks to reach those Blue Oceans and Unconsumers. "Fermentation" is the alternative.