Korva is a scroll of the internet’s best ideas.
Lab Priorities
The big labs' priorities are easy to lay out. Take the goal at OpenAI, Anthropic, and DeepMind to be AGI or ASI in a safe way while capturing a lot of profit. The priorities that ladder into that probably look like ChatGPT, ads, and coding, with an expansion after that such as Cowork, which gets different types of users doing richer tasks across more interfaces.
Judging Competitiveness
The useful work is judging how competitive each of those efforts actually is and how far each one reasonably reaches, then holding that up against each company or opportunity on the table. Circular logic about what people believe of the big-lab strategy, and how any application can live, teaches little for a decision.
The Next 99%
A grand strategic framework about which layer wins takes up a lot of investing energy. Spent instead on the next 99% of diffusion, that energy goes to how far the technology still has to spread if the world is only 1% of the way in.
How fast have past technologies spread?
Economist Paul David examined the slow spread of electricity in a 1990 paper called The Dynamo and the Computer. Factories had electric motors for decades before productivity gains appeared, because they had to redesign floor layouts around them. Steam-era factories used one large engine with belts and shafts, while electric motors let each machine run on its own. The full benefits arrived in the 1920s, roughly forty years after the first power stations opened.
Why does value shift between technology layers?
Harvard's Clayton Christensen described the law of conservation of attractive profits, in which profit moves to whichever layer is hardest to substitute as other layers become commodities. In personal computers, Intel chips and Microsoft software captured most profits, while makers of the boxes competed on thin margins. The pattern earned the nickname Wintel. In earlier decades, IBM had held the profits as the dominant seller of complete systems.
More from Invest Like the Best with Patrick O'Shaughnessy
