For The Better - Email 8/22

For the Better comes to you bi-weekly with ideas about how and why to build companies focused on human flourishing and stories of the people who are doing it. Other enthusiasms may occasionally appear.

AI and Corporate CultureWant to take a guess on the number of books about management practices and their various pros and cons? Me neither. For the purposes of this post, the one relevant fact about their existence is that they’re all based on quantitative and macroeconomic analysis along with case studies.From Good to Great and Built to Last all the way back Peter Drucker’s The Practice of Management which came out in 1954, they’re alike in that sense even as each of them reflects the business context in which it was written. (This is a nicely varied list of ten titles that span the decades if you’re inclined to do a deeper dive). Now, AI is giving us a whole new way to study the humans who work in organizations, as “What if AI had corporate culture?” shows in fascinating detail. The premise of the experiment is simple, and the results are really compelling:“Multi-agents are taking over. Research shows that AI systems with 30+ agents out-performs a simple LLM call in practically any task (see More Agents Is All You Need), reducing hallucinations and improving accuracy.But how should agents actually work with each other? While exploring methods to improve AI performance on software engineering tasks, I stumbled upon an idea: What if we structured AI agent interactions like the org charts of big tech companies?Here are some key takeaways I learned after organizing groups of AI agents as if they were in companies like Apple, Microsoft, Google, Amazon, and more:

1. Companies with multiple “competing” teams (i.e. competing to produce the best final product) like Microsoft and Apple outperformed centralized hierarchies.

2. Systems with single points of failure (for example, one leader making important decisions) like Google, Amazon, and Oracle underperformed.

3. Big-tech organizational structures had a modest but noticeable impact on problem-solving capability.”

Fundamentally, corporations are just slow AIs, so the lessons of management and governance we have are really useful when it comes to thinking about AI – and I’m confident they’re only going to become more useful.Multi-agent coordination is the same puzzle regardless of the kind of agent. AI, in turn, can give us a really interesting perspective on human behavior. Of course, since no actual humans are involved, its conclusions may not be totally correct, but that doesn’t mean we can’t learn a ton anyway. Think of it as the 21st-century version of the Wright Brothers and their wind tunnel. Without it, they might never have gotten off the ground at the turn of the last century.

Things I’ve Enjoyed Lately 

🔵 Uncovered Euripedes fragments are “kind of a big deal”“Actually, it has a relatively happy ending. It’s not one of these tragedies where everyone winds up dead.” What else do you need to know to click on this article about the contents of the recently discovered papyrus sent by an archeologist in Egypt to two classics scholars at the University of Colorado? 🔵 The secrets of out-performing family businessesCommitment to purpose and a long-term view are just two of the shared characteristics called out in this report. Rigorous processes for hiring talent and investing are also on the list. Imagine a world in which the practices of family-owned companies like these were commonplace for all companies.

Athena:

One of the things I enjoy the most about what I do is that every day is a little different. I meet people building incredible companies and doing great things for the world, do my writing and work, and always reserve time for my family. Athena, which offers executive assistants at affordable rates, shares my belief that the key to making the most of everything life has to offer is keeping it running smoothly. I've partnered with them so you can focus on living your life rather than managing it. 

More info and a discount offer are here.

🔵 The Platonic Representation HypothesisThis paper makes an argument that as AI models align, the world is being brought closer and closer to “a shared statistical model of reality, akin to Plato's concept of an ideal reality” – a so-called “platonic representation.”