Week 38 · 2025-09-15 → 2025-09-21 · 17 newsletters

Reps, Wealth, and Quiet Weeks

ai-and-the-craft-question · wealth-philosophy · career-and-the-stage · interior-weather

A genuinely thin week. Seventeen emails across seven days, no breaking news thread, no dominant industry story. What did show up were four loose clusters: a small AI conversation that kept circling the same craft question, two pieces on wealth as a philosophical problem rather than a financial one, a career-and-stagecraft set worth bookmarking, and a handful of personal essays that were the strongest writing of the week. Treating this as a five-theme week would be padding. It was a four-theme week, and the interior writing carried it.

AI and the Craft Question: Reduce Friction, Or Skip the Reps?

The week's AI conversation never reached the agent-economy fever pitch you saw in adjacent weeks. What it did do, in a quieter way, was return to the same uncomfortable question from three angles.

Amanda Natividad at The Menu wrote the cleanest version in "AI can save you time, but don't let it steal your reps." Her working theory: AI reduces friction, and friction is where craft is forged. The argument is autobiographical. She came up on Twitter writing thousands of often-bad posts that sharpened her instincts for what lands. If LLMs had been what they are today, she is honest that she might have skipped that ramp and ended up "indistinguishable, passable." The prescription is to use AI to reduce just enough friction (drafting a note to your kid's teacher, summarizing a staff meeting) but not so much that you skip the reps that compound.

Sean Ellis ran the inverse argument in "AI as the New Mentor," that for new marketing grads AI may be the best thing that ever happened to them, and that the smartest founders are already taking advantage of it. The two pieces sit in direct tension. Natividad's frame is that AI removes the productive struggle that builds taste. Ellis's frame is that AI removes the gatekeeping that prevented juniors from operating at senior leverage. Both are right in different domains.

Carilu Dietrich widened the lens in "Agents Are Buying, Agents Are Selling," her preview of the Agentic Distribution Summit. The question she asks (what does it mean when agents are not just assisting but actually buying and selling, sometimes only talking to other agents) is the organizational version of Natividad's craft question. If the agent does the rep, who builds the taste?

The single market-level read came from Citrini Research with a single-stock long thesis tracing the AI infrastructure progression from servers (Nvidia, Supermicro, Celestica) to data center interconnect (Credo, Ciena) to high-voltage power (American Superconductor, Siemens Energy) and now to the new data centers themselves. It is the most disciplined map of the buildout I have read in a while, and it makes the implicit point that the trillion-dollar capex story has at least three more chapters in it.

The take: the operator-level question this week is not whether agents work. It is whether the humans operating them are still doing the reps that produce taste. Natividad is right that the temptation to skip the bootcamp is the actual risk, and Ellis is right that the bootcamp itself is changing. Both have to be true at once.

Wealth Philosophy: Two Pieces, Same Argument

JJ Chou ran two posts back-to-back that, read together, form one of the most interesting wealth arguments of the week. The first was his fourth installment on The Wealth Ladder, built around the Kahneman-Deaton 2010 finding that more money does correlate with feeling better about life, but only up to about $75K (roughly NT$2.25M), after which additional income does not buy additional happiness. His reframe: at Level 1 or Level 2 of the wealth ladder, more money genuinely does buy more happiness. The threshold matters more than the absolute number.

The second post, "Buy-Borrow-Die," walked through the wealth-preservation strategy that ultra-high-net-worth Americans use legally: buy appreciating assets (stocks, real estate, equity), borrow against them rather than selling (no tax on borrowed money, interest sometimes deductible), and pass them on at death with a stepped-up cost basis that erases the capital gains liability. Chou's treatment is plainspoken and treats it as what it is: a feature of the tax code, not a loophole.

Investing Insights ran the supporting piece, "10 Things I've Learned About Wealth Management in 10 Years," with the line worth keeping: "the only benchmark that truly matters is whether you are on track to reach your goals." The companion observation that no number truly makes you feel wealthy because it all depends on expectations and surroundings is the same point Chou is making from the income side.

The take: the wealth conversation this week converged on a single uncomfortable truth. Both the income ladder and the asset-preservation playbook are governed less by what you have than by what you compare it to, and the people who handle wealth well are the ones who fight the comparison reflex instead of feeding it.

Career and the Stage: Job Doubt and Public Speaking

Ami Vora at The Hard Parts of Growth wrote the most actionable piece of the week on the limbo of considering a job change without actually moving. Her prescription is an 8-week focused window: set a timeline, do real research on what is out there, and decide. The insight that the half-in, half-out state is the most insidious failure mode (you stop being energized in the current role, but you also do not actually look) is the one I will hold onto.

Vince Pierri at Growth Unhinged (guest posting for Kyle Poyar) ran a long, useful framework for the better keynote, built around a three-move intro formula: pain point, then promise, then proof. The 300-talks experience shows. The line that landed: the number one obstacle is not stage fright, it is structure, and when leaders do not know where their talk is going their nerves spike because their outline cannot hold them up.

Rob Thomas at The Mentor ran the short pep-talk version in his Aspiration installment, anchored on David Schwartz's claim that belief sets the upper bound on achievement.

The take: Vora and Pierri are the keepers. Both pieces are bookmarkable and both attack the same underlying failure mode, which is operating without a structure that forces you to commit to a take. The job-search version and the keynote version are the same problem in different rooms.

Interior Weather: The Best Writing of the Week

Three personal essays were the strongest writing of the week, and they had nothing to do with industry signal.

Steven Schlafman at Where the Road Bends wrote "Fighting Ghosts," a slow, scary account of a strange abdominal sensation that surfaced during a J.S. Ondara concert at Levon Helm Studios and would not leave. The piece is about the stories we tell ourselves to avoid panic and the moment those stories stop working.

Abby Falik at Taking Flight wrote about a Bodhisattva retreat at a California redwoods monastery, the invitation to give a short talk that came the night before, and the response when she asked how to prepare: "the only way to prepare is not to." The conditioned self that performs and proves vanished. What replaced it she calls flow, trust, aliveness.

Justin Mares at The Next traced the 1997 FDA guidance document that quietly made the modern pharma ad possible, which is the kind of regulatory archaeology that explains more about our information environment than most policy posts do. Alec McNayr ran an open letter to the freeway off-ramp line-cutter that was the funniest thing in the inbox this week, and Lingthusiasm interviewed Pius Akumbu on tone in Babanki and the linguist's project to found a primary school in his home village. None of these were trying to predict anything. All of them were the better reads.

The take: in a sparse industry week, the personal essays do the work that the trend pieces cannot. Schlafman and Falik are the kind of writing you save and reread, and the lesson for anyone who writes a newsletter is that the weeks when you have nothing industry-shaped to say are the weeks the interior writing earns its slot.

Three Takeaways from the Week

The AI conversation this week was a craft conversation in disguise. Natividad and Ellis pulled in opposite directions on the same question, which is whether the next generation of operators will still get the reps that build taste. The answer is going to be domain-specific and the operator-level call is going to be how much friction you intentionally preserve in your own practice. Carry that question into the next quarter.

The wealth pieces from JJ Chou and the Investing Insights wrap-up converged on a single point worth holding: wealth is governed by expectations more than by numbers, and the playbooks that look like loopholes (Buy-Borrow-Die) are mostly just disciplined applications of the rules as written. The interesting question is not how to copy the playbook but how to set the expectations that make any of it feel like enough.

If you only revisit three pieces from the week, I would suggest Amanda Natividad's "AI can save you time, but don't let it steal your reps" for the cleanest framing of the AI craft question, Steven Schlafman's "Fighting Ghosts" for the personal essay that hits hardest, and Abby Falik's "Will we choose love, or fear?" for the post most worth reading slowly in a week that does not demand speed. Sparse weeks reward the writers who knew what kind of week it was.