Cognitive Revolutions

The board revolt at OpenAI reminds me of my favorite idea from Yuval Noah Harari’s Sapiens.

The idea of the imagined intersubjective order. Companies are concepts of our collective imaginations. We call them legal entities and assign liability to them. They can be a vehicle for capturing value created by humans. They receive tax assessments. They can even be sued — absolving the people in charge of direct responsibility.

Yet, a company can be dissolved and reincarnated elsewhere by having the employees simply say they resigned and have moved to a new company. And that's how we keep lawyers employed. Tricks of our own cognition.

I've probably repeated this one enough.

Excerpt: “Companies are legal fictions which we find convenient to use to move capital around and balance accounting ledgers.“

It's taken a long while. But I see it now. Store information for convenience. The search bar exists… and generally works

A good passage in continuation of the note on collective imaginations.

Responsibility that should be borne by an individual is dissolved amongst a group, who rely on the process of deliberation, passing on the responsibility to the organization.

Excerpt: “Committees are commonly used in our society because they create the illusion of avoiding risk. They are a wonderful device for avoiding responsibility while making the institution seem more rather than less accountable. Modern institutions have overloaded on actual risk while fleeing the appearance of it, especially if you count “failing at core mission” as a risk. Such aversion to the appearance of the unusual can’t be justified on economic grounds. Rather, it is a socially driven aversion.“

Technological change is only possible with societal (behavioral) change

To use databases, you need to be able to think about your data in a structured way. The people who could do that gained most of the leverage from using databases. The folks who made it easy for people to transition to databases (ERPs and Co) captured the rest.

AI systems are fundamentally scalable decision making engines. And data is the oil required to power it. If you want to properly leverage AI, you need to make sure you're fueling it with the best refined data.

Otherwise, you're no different than the drunk trying to reliably operate complex machinery. The only repeatable part of that is injury.