Reasoning from first principles is the most popular approach to problem-solving right now. It means breaking a problem down into its basic components. Then looking for constraints and solutions based on those.

The most famous proponent, and one of the main reasons for its popularity, is Elon Musk. Given his successes with multiple technology companies over the years, I think the popularity is well deserved.

However, I think reasoning by analogy is just as important. In my experience, a lot of the problems we encounter, whether it be trying to explore new product insights or leading a family, are complex in ways that are hard to reduce to first principles.

In these cases, it’s better to expose yourself to lots of instances; get lots of reps. You can learn by doing or you can study cases from history. This allows you to develop a feel for what can happen, and how a situation might play out.

The bigger the database of experiences you have, the better your intuition will be at navigating a novel instance of a problem.

And if you find that your internal Elon is arguing against reasoning by analogy already, pass the mic to one of history’s best investors to debate him. Charlie Munger is no slouch.

Note: if you’re interested in digging deeper into these approaches, I recommend reading Range by David Epstein and Peak by Anders Ericsson. You can also ask your favorite chatbot about Cognitive Flexibility Theory and Cognitive Transformation Theory. And If you’re feeling ambitious, check out the book Accelerated Expertise.

First you will come to the Sirens who enchant all who come near them. If any one unwarily draws in too close and hears the singing of the Sirens, his wife and children will never welcome him home again, for they sit in a green field and warble him to death with the sweetness of their song.

The Odyssey, Book XII

I watched a documentary about the creation of index funds. It’s called Tune Out the Noise and was produced by Dimensional Fund Advisors. Along with Vanguard, Dimensional is one of the pioneers of passive investing.

The basic premise of passive investing is that markets are very hard to beat. In other words, you’re better off investing in a diversified market index fund. Trying to pick a handful of stocks that you think will outperform is a fool's errand.

It’s worth noting that Dimensional’s original team includes most of the biggest legends of modern finance (Eugene Fama, Ken French, David Booth, Rex Sinquefield, John McQuownm and later Myron Scholes, and Robert Merton). These guys effectively invented modern financial analysis and coined the efficient market hypothesis.

Meanwhile, the Youtube ads algorithm can’t help but flash ads selling financial services. Some of these are particularly cringe advertising day trading and special tools to beat the market.

I found that a little amusing.

Agents are clearly going to have a big impact on work. They will probably have a big impact on our daily lives.

I can picture lots of ways that I would leverage a reliable agent throughout the day. In fact, I think in the near future, I’ll have at least a couple of agents working on something at all times.

The keyword is reliable.

How will I make sure the agent isn’t making mistakes on my behalf along the way? Or even better, how do I make sure that the mistakes the agent makes are trivial to fix?

I wonder how I will manage all these agents. A kanban board probably will be too simplistic. I’m probably going to need something more sophisticated.

I wonder if I will use multiple agents built by different developers the way I use different chatbots like ChatGPT, Claude, Grok and Perplexity.

If that’s the case, will they be able to interact somehow?

Will I have to manage them all separately across different interfaces?

I sure hope not.

I wonder what the agent operating system will look like.