The Interesting Times: Living with Your AI Shadow

I talked with my friend Brian David Crane on his podcast this month about the tradeoffs between assuming agency and accepting what you can’t control, how to use AI in a museum, and my philosophy of history. You can watch it on YouTube or listen on any popular audio streaming platform.

I'm also speaking on a panel tonight (Tuesday, the 29th) in Austin about Decision-making in Turbulent Times. It’s free, and there are a few spots left if you’d like to come out.

This issue is a bit longer issue than normal as I found lots of good stuff this month. Sections added for easy skimming - AI, psychology, and markets being the main topics. As always, thoughts and feedback are welcome.

Department of AI

Autoamputation Flow [Blog]

Contraptions

"As Marshall McLuhan once observed, "Every extension is also an amputation." The wheel extends the foot, and in doing so, removes the necessity of walking. The book extends memory, and in doing so, weakens the habit of remembering. With AI, what is extended is the head — the seat of thought, language, judgment, and becoming. And what is amputated is not merely a capacity but a process: the recursive unfolding of the self."

Autoamputation flow is a term to explain the continuous process by which we externalize parts of our cognition into AI, only to receive back not mirrors of our current selves, but "a kind of delayed, distorted continuation of what we were. The AI, drawing from our past actions and the aggregated tendencies of others, feeds back a synthetic version of "us," which then shapes our next step forward."

We are still relatively early in understanding how interacting with AI will shape us, but, ultimately, we have the path to either blend in with it and rest or to begin again and reintroduce novelty into the system.

We have to maintain some meaningful difference between our cognitive patterns and those of our AI extensions or fall into some sort of cognitive entropy where we merge with the patterns of our digital shadows.

Care, then, becomes a refusal to coast. It is the act of gently outrunning your own average, keeping the delta alive not through panic or performance, but through quiet acts of novelty and noticing. It is living with the awareness that your AI shadow is always trailing you — and loving it anyway.

We distort our own patterns not to disappear, but to contribute something slightly stranger — and slightly more human — to the pool.

Revenge of the Junior Developer [Blog]

Sourcegraph

Thoughts on how AI will impact developers and how product and engineering teams will shift going forward from Steve Yegge. Yegge wrote something of a cult classic piece in the 2000s, comparing Google and Amazon’s business structure - which has aged quite well.

The most common argument today is that AI will mostly replace junior developers since senior developers with more architectural experience will be able to use AI agents to do most of the "grunt work."

Yegge argues that junior developers are actually adopting AI more quickly and may actually see something more like an innovator's dilemma situation on the career level, where junior developers leapfrog more senior developers.

He also gives a nice history of the waves of AI agents from chat-based to agents to what he calls agent clusters, a model where developers orchestrate multiple AI agents working on different tasks simultaneously. He argues companies should be spending tens of thousands of dollars per year per developer going forward on AI compute to optimize their engineering function.

The new job of a software developer going forward will soon be managing dashboards of coding agents and their AI supervisors… Some might derisively call this job *babysitting*, and accuse AIs of being little whiny baby robots that need a grownup to cut their food into little pieces, and change their diapers, and clean up their messes, and keep them from wandering out of their playpens. But we prefer to call it *software development*. This is our destiny.

Department of Psychology

Unpacking Trauma: How Early Wounds Shape Behavior and the Path toward Healing | Jeff English [Podcast]

Peter Attia

I have a lot of trouble with the word trauma because it seems to be used to cover such a wide range of people's experiences, from living through war or serious family violence to something as small as a perceived slight.

I thought this podcast did a nice job of unpacking the spectrum there and having a thoughtful conversation about how events of any magnitude can impact people and ways for thinking about and addressing them. The concept of the trauma tree as a metaphor for thinking about the different ways in which we can talk about types of trauma was useful.

The roots of the tree include abuse (physical, emotional, social), neglect (something that failed to happen for me), enmeshment (boundary violations, emotional incest), abandonment (physical or emotional), and tragic events.

These wounds, especially during formative years, can cement themselves and affect development throughout life.

Jeff illustrates this with the story of a four-year-old boy who uses deception (pretending to vomit) to distract his father from beating his mother. The boy learns, at some subconscious level, that manipulation and deception are effective strategies for protecting people he loves and himself.

You can look at this as a very effective strategy for the situation in which he found himself, but you can also then see that an inability to recognize that pattern and ‘acting it out’ subconsciously could be harmful to future relationships.

What is Stage Five (like)? [Video/Article]
Meaningness

The idea of psychological stages has been used by a few psychologists, most notably by Robert Kegan in his model of adult cognitive, affective, and social development. Though Keegan himself avoided naming the stages, David Chapman has called stage four "rationality" and stage five "meta-rationality" and talked at some length about the difficulty of bridging stage four and stage five.

The Stage 4 mind seeks to understand the world through systems and frameworks—the kind of thinking that builds spreadsheet models and five-year plans. It's the mode where expertise is defined by mastery of specific methodologies.

The meta-rational Stage 5 mind does something fundamentally different. Rather than being constrained by methodological purity, it recognizes that elements from different systems "can be deleted, combined, edited and remixed in ways that more accurately model reality, regardless of whether the resulting system is ideologically or logically consistent."

I've enjoyed his writing and thinking on this topic, and found this description of what operating in stage five relative to stage four feels like a helpful way to try and talk about ot.

I find myself often reflecting on the same metaphor he ended up using: gardening.

[Stage 5 is] more like tending a garden than like building and operating a machine, which is the experience of stage four. It’s more nurturing, less controlling. At stage four, you relate to everything in terms of "What does this mean *to me*? What do *you* mean to me? What can *I* do with this thing?"

Although, a garden is still pretty top down; like, you decide where to put which rose bush, and you put some tulips over *here*. Maybe a better metaphor would be **taking responsibility for a plot of woodland** that you nurture. So you make sure that there’s adequate water in a drought. You clear out diseased trees. You build brush piles to provide habitat for small mammals. Foresters do this. They pile up dead branches, and rabbits or weasels, or I don’t know what, live in there.

This metaphor of "nurturing" might sound *nice*. And that’s not really the point. Part of caring for a plot of woodland is uprooting invasive plant species. It’s setting traps for pest animal species. It’s building a fence around the plot to keep out wild dogs. If 30 to 50 feral hogs break through the fence, a semi-automatic rifle might be called for.

Dr. Becky Kennedy — Parenting Strategies for Raising Resilient Kids, Plus Word-for-Word Scripts for Repairing Relationships, Setting Boundaries, and More [Podcast]
Tim Ferris

I am a new parent and I do not know what I am doing! I have thoughts and feelings about how I was raised and how I see other people raising their kids, and I felt like this podcast put words and structure around a lot of those thoughts.

Dr. Becky's approach to parenting is that you should start with the belief that you and your child are fundamentally good people at your core - even when you're struggling. This perspective shift changes everything about how we respond to challenging behavior, moving us from judgment to curiosity.

Most parenting advice operates on a first-order level: child misbehaves → parent corrects → behavior changes. But this very linear, Pavlovian model fails to account for the complex emotional dynamics and underlying causes that drive behavior. People are not laboratory mice, understanding where the behavior comes from and where the correction comes from matter.

Dr. Becky offers a more sophisticated model through what she calls "sturdy leadership," which combines "equal parts very firm boundaries and parental authority as it does warm, validating connection." This balance creates psychological safety while maintaining clear expectations.

Department of Markets

Michael Sonnenfeldt. Post-Liquidity Ultimate Wisdom [Podcast]
Exit Paradox

A thoughtful podcast with the founder of Tiger21 (UHNW group for people with $20 mm+). It was pretty interesting to hear about the things he's observed across members in the group for the last 25 years or so. Some points that jumped out to me…

On why lots of entrepreneurs end up being bad investors:

"Most entrepreneurs focus on a single opportunity. As an investor, if you do that, you'll be out of business. Because once you start investing passive capital, if you're not prudently diversified, then any one mistake could destroy your portfolio.

One of the worst training grounds for being a competent investor is being a successful entrepreneur.

…

when you have an underlying business that can make profits this year and next year and the following year, it allows you to become a sloppy investor because you can make investments that you lose money on and you don't think much about it because the next year's profits replenish the wealth, so to speak."

On why you should take 3-5 years after a large windfall to invest the money:

"Rather than making a sale in January and being fully invested by December, the best advice I would give is not precise, but take three to five years before you are fully invested. Because you just don't know what you don't know.

So one thing is time. And for sure, the other thing is risk. Most entrepreneurs were successful in part because they found a sweet spot that married their skills and some need in the world. Investors really don't have sweet spots. And entrepreneurs who become investors don't realize how ill-equipped they are to be dispassionate, diversified investors."

On parenting:

In the end, most kids want contact with their parents.

…

At any economic level, those children that have loving, involved parents on average will do better than those who do not.

I remember someone said ‘I'm not giving my kids anything.’ And I thought, oh my god, here comes another story about how hard it was for me to earn money and I want to make it just as hard for my kids because otherwise they won't ever fulfill themselves.

But as I was thinking about that in my mind, he said, ‘no, no, you don't understand.’ I'm not gonna give my kids anything, but I'll invest everything in them. And I think while it's not a perfect way to think about it, the notion of investing in children's futures [is pretty helpful].

Why Patience Matters During Market Stress [Article]
Man Group

Trend following is a simple, if counterintuitive, investing strategy: buy things that are already going up and sell things that are already going down. This is the opposite of what most people think is smart investing advice! However, there's pretty robust evidence over many different markets and hundreds of years showing that trend following strategies work.

They are particularly interesting for investors with a lot of stock and bond exposure, as they're historically uncorrelated to those making it a valuable addition to stock-bond focused portfolios.

It's kind of a weird thing as to why trend following works. The most plausible explanation to me is that there are lots of periods of time where it kind of sucks, and one of those periods of time is this year where you have these sharp sell-offs that then rebound and create a whip-saw that's really bad for trend investors. No pain, no premium, as the saying goes.

This gives a nice overview of trend following strategies in major market crashes, showing that the first 10-15% down in a market tends to be bad for trend following strategies and they tend to kick in as the market goes further down.

Heliocentrism: Objects may be further away than they appear [Article]
J.P. Morgan

One of the segments of the market that I understand least is the energy market and it’s also one where I hear wildly different interpretations depending on who I am talking with.

This report shows a disconnect between the breathless narratives about solar's inevitable dominance and the more sobering reality of global energy systems.

Despite $9 trillion spent globally over the past decade on wind, solar, EVs, energy storage and grid infrastructure, the renewable transition is moving at a glacial pace of just 0.3%-0.6% annual increase in renewable share of final energy consumption. Solar capacity may be booming worldwide, but the transition is proving stubbornly linear rather than exponential.

By contrast, the shift away from open hearth furnace steel production began in 1960 and was mostly done by 1980. Why the big difference?

At the time, new steel technology reduced production times and energy use by 80%-90%. In other words, the steel transition paid for itself and rewarded early adopters. That’s not the case with renewables.

Most renewable technologies do not entail massive improvements in cost or energy efficiency. They can sharply reduce carbon footprints, but that does not create substantial economic incentives in a world without a large universal price on carbon. Fast historical industrial transitions financed themselves via returns accruing to innovators and early adopters, while the gradual and linear renewable transition shown on the prior page has required $9 trillion of global spending since 2010 to move along.

Today, solar power accounts for ~7% of global electricity generation. Electricity typically accounts for under one-third of final energy consumption in most countries. Electricity is primarily used for space cooling, refrigeration, computers, and other electronic equipment, and a small share of space heating in commercial/residential buildings. Transport, industrial production, and most building space heating still rely on fossil fuels. As a result, when you boil it all down, solar power accounts for ~2% of global final energy consumption

At the same time, ice core analyses show CO2 levels close to 270 parts per million (ppm) by volume during the preindustrial era; in 1958 when Mauna Loa monitoring began they reached 313 ppm; by the year 2000 they were 370 ppm and by the end of 2023 they reached 420 ppm, more than 50% above the late 18th-century level. This rise, together with contributions by methane and nitrous oxide, has translated to about 1⁰C of global warming compared to the 19th-century mean. All continents have been affected, recent decadal warming gains have been steadily rising, and the eight years between 2015 and 2022 were the warmest years on record.

Nuclear seems like the logical way out of this, but it seems stuck in a political purgatory where it is not "really green" and, of course, Chernobyl. Or more robust carbon pricing? Does that work in a global world?

Here are a few more things you might find interesting:

Interesting Essays: Read my best, free essays on topics like bitcoin, investing, decision making and marketing.

Consulting & Advising: Are you looking for help with making decisions around scaling your company from $500k to $5 million? I’ve been working with authors, entrepreneurs, and startups for half a decade to help them get more out of their businesses.

Internet Business Toolkit: An exhaustive list of all the online tools I use to be more productive.