Weekly notesNo. 01

Bets, flashlights, and growing up

Thinking in Bets

I’m only a third of the way in and it’s already been added to the bookshelf. Every decision we make is a bet. We can become better decision makers by training ourselves to think in bets, and poker might be the ideal training ground.

“Thinking in bets starts with recognizing that there are exactly two things that determine how our lives turn out: the quality of our decisions and luck.”

More passages I marked4 quotes

“Take a moment to imagine your best decision in the last year. Now take a moment to imagine your worst decision. I’m willing to bet that your best decision preceded a good result and the worst decision preceded a bad result. Drawing an overly tight relationship between results and decision quality affects our decisions every day, potentially with far-reaching, catastrophic consequences. I have yet to come across someone who doesn’t identify their best and worst results rather than their best and worst decisions.”

“Chess is not a game. Chess is a well-defined form of computation. Trouble follows when we treat life decisions as if they were chess decisions. Chess contains no hidden information and very little luck. If you lose at a game of chess, it must be because there were better moves that you didn’t make or didn’t see. Chess, for all its strategic complexity, isn’t a great model for decision-making in life, where most of our decisions involve hidden information and a much greater influence of luck. Poker, in contrast, is a game of incomplete information. It is a game of decision-making under conditions of uncertainty over time.”

“We are discouraged from saying ‘I don’t know’ or ‘I’m not sure.’ We regard those expressions as vague, unhelpful, and even evasive. But getting comfortable with ‘I’m not sure’ is a vital step to being a better decision-maker. We have to make peace with not knowing. Admitting that we don’t know has an undeservedly bad reputation. There is a great quote from physicist James Clerk Maxwell: ‘Thoroughly conscious ignorance is the prelude to every real advance in science.’ I would add that this is a prelude to every great decision that has ever been made. What makes a decision great is not that it has a great outcome. A great decision is the result of a good process, and that process must include an attempt to accurately represent our own state of knowledge. That state of knowledge, in turn, is some variation of ‘I’m not sure.’”

“What good poker players and good decision-makers have in common is their comfort with the world being an uncertain and unpredictable place. They understand that they can almost never know exactly how something will turn out. They embrace that uncertainty and, instead of focusing on being sure, they try to figure out how unsure they are, making their best guess at the chances that different outcomes will occur. The accuracy of those guesses will depend on how much information they have and how experienced they are at making such guesses. This is part of the basis of all bets.”

Protocols

It’s a massive book. I’m not sure how long it’s going to take me to get through it, but I’m eager to try. I respect Huberman’s curiosity and devotion to optimizing body and mind.

“True optimization is about rescuing time, not just increasing the number of years of your life, but by increasing the number of effective hours of your days (and the number of hours of great sleep at night).”

More passages I marked2 quotes

“Better living through chemistry still requires better living.”

“Individual experiences are often dismissed because academic science does not generally consider a sample size of one person as evidence (except in formal case studies, of course). And yet, each of our lives is its own kind of laboratory. There’s nothing wrong with giving credence to your own experience and to watching how something affects you as a source of data.”

Huberman calls this “anecdata.” I’ll be stealing that.

AI is now helping build the next AI

Anthropic’s pace of AI development is accelerating, estimated at about twice as fast as a year ago. Most of that is attributed to using AI models to improve the next models.

Why that worries me, and why I'm conflicted

The estimate comes from METR (Model Evaluation & Threat Research), an independent nonprofit that tests frontier AI models for dangerous capabilities. They work closely with the big labs but don’t work for them, which is why their numbers carry weight. In a report published on 5 November, they estimated that Anthropic is now improving its models at roughly twice the pace it managed between early 2024 and late 2025 (somewhere between 1.3x and 3x). They’re confident the jump isn’t just more computing power, and they put most of it down to Anthropic using its own AI to do AI research. One caveat: Anthropic redacted much of the underlying evidence, so outsiders can’t fully check the work.

This is the recursive self-improvement that everyone taking AI safety seriously is scared of. As AI models become more central to training new AI models, humans are increasingly left out of the loop. In theory, we may eventually fall out of it completely, leaving us with little to no insight into, or control over, the models’ growing capabilities, or whether they’re aligned with human wants and wellbeing.

I have very little insight into the realistic odds of catastrophe here. And I’m a little conflicted. I’m an AI power user, and I’m excited to see models become ever more powerful, capable, intelligent, and useful. But I can also clearly see how recursive self-improvement leads to the birth of a mind that’s more intelligent than us. That’s equal parts exciting and scary.

Exciting

Superhuman intelligence could answer many of our deepest questions and solve many of our most pressing problems. Selfishly, I also want my own Jarvis from Iron Man.

Scary

Among other things, superintelligent minds may see us as uninteresting as ants, or worse, as obstacles to remove. If it can use its intelligence to keep improving itself, that quickly leads to something beyond our comprehension.

I see no reason to think we would be able to reliably control something smarter than us.

Why “prepping” makes people so emotional

This past week I’ve been thinking about prepping (for a disaster, or the “apocalypse”) more than usual. Maybe it’s the escalating situation in Iran, the acceleration of AI (particularly its cybersecurity capabilities), or governments around the world urging their citizens to stockpile food and water. Probably some combination.

Because it’s been on my mind, I’ve had a lot of conversations about it with family and friends, and noticed that the topic stirs up strong emotions whatever your stance. I haven’t worked out exactly why yet, so this week is mostly observations, including about myself.

My view on prepping is tied to a bigger set of ideas I’m writing about. More on these in coming weeks:

  • Risk & rewardFamiliar to traders and investors, but it applies to most of life.Coming soon
  • InsuranceA valuable tool rather than a burden, and it takes many forms beyond a classic policy.Coming soon
  • Thinking in betsAlmost every decision is made under uncertainty.See above ↑

Observation 1“Prepping” covers wildly different things, so we talk past each other.

  1. Short outagehours to a day
  2. Temporary disruptiona few days
  3. Extended outageweeks
  4. Major disruptionmonths
  5. Long-term scenarioa year or more
Read the observation

Prepping as a blanket category isn’t very helpful. Someone prepping for a two-day power outage by buying candles and flashlights is not the same as someone prepping for a year-long apocalypse.

I think this is one of the first places the conversation gets emotional. One person brings up prepping (wherever on the spectrum they mean) and the other assumes they’re suggesting we’re days away from The Walking Dead. It’s easy to talk past each other and end up seeing the other person as ridiculous.

It isn’t always easy to make sure we’re talking about the same thing, because we don’t always know what we’re talking about. None of us knows what the future holds, and there are so many competing scenarios, each with its own (largely made up) probability.

A good place to start is probably a rational conversation about the spectrum itself. What might each point on it actually look like, and how well or poorly would you do in that scenario today? Then, from a place of curiosity and respect for uncertainty, compare each person’s estimated probability for each scenario, with their reasons.

That might still end in frustration, with each of you convinced the other is confused. Or it might help you land on an agreed, sane level of preparation, whether that’s flashlights, food and water for some number of days, or something more extreme. To each their own.

Observation 2Everyone is already overloaded.

Read the observation

People are already mentally stretched. Finances, relationships, kids, the ever-expanding admin of life: I get the sense that everyone is exhausted. “I’ve got enough things to worry about. Shut up about imaginary scenarios.”

The emotions come in when you’re overworked and already stressed, and it feels like someone is making that worse by filling your head with their own fears about situations that probably won’t matter.

I see this more as a critique of modern society and the growing demands, pressure, and admin it puts on us. I’m not saying we should simplify our lives so we can spend the freed-up energy planning for every way society could crumble. I’m pointing at a way of living so exhausting that we may have no choice but to ignore real risks to the safety and security of our families.

Next week: more on what I actually think about prepping itself.

What does it mean to be mature?

We all use the words “mature” and “immature,” but I don’t have a precise definition for either. The more I look into it, the more complicated it gets.

I recently heard Jordan Peterson argue that you’re never fully mature until you have children, because for most people a child’s birth is the first time they care about something more than themselves. To defend that, he first had to define maturity:

“The more forward looking you are, the more mature you are. The more your actions in the present are constrained by the future, the more mature you are.”

Two kinds of immaturity, and why it matters

I like this definition. It suggests immaturity comes in two forms:

  1. Ignorance. Whether willful or not, you fail to notice that today’s choices have long-term consequences, so you drift.
  2. Inaction. You see the consequences clearly, but fail to overcome whatever resistance is associated with dealing with them. Maybe out of procrastination, because the short-term pleasure wins, or because you don’t feel much connection to the person you’ll be in five years.

I find the second more common. Most of us aren’t ignorant of what we should do. We just tell ourselves we’ll deal with it later.

There’s a common assumption that maturity means being serious, boring, and unable to have fun. I don’t think that’s right. There’s a difference between taking yourself seriously and taking serious things seriously. You can be silly, have fun, and still give consequential things the weight they deserve. I’d argue that’s the goal. Growing up shouldn’t mean losing play, because then we risk avoiding growing up altogether.

I’d argue that avoiding maturity underlies a lot of the world’s problems. We need mature adults to keep order and protect children. Far too many of us are blind to things with long-term consequences for ourselves, our families, our communities, our countries, and the planet. Even more of us see them and still fail to act.

So maybe we need a more nuanced definition: one that leaves room for silliness and play but still asks us to get our act together. Defined that way, maturity might stop being something we resist and become something we aspire to.

My working definition

Maturity is noticing what needs to be dealt with, and dealing with it before it gets worse, without forgetting how to have fun along the way.