Pirsig's concept of dynamic quality is what the best value investors actually hunt. This has been the most useful frame I've picked up in years.
Buffett went from cigar butts to wonderful businesses at fair prices. Same word, different math. The job is finding the next layer before it becomes the consensus screen.
Spent an hour with @bogumil_nyc on this and how Top Mark Capital applies it.
TALKING BILLIONS: @mikenicoletti : Sailboats & Telltales, Roads & Motorcycles, Value & Quality, Good People & Generous Mentors — From Jacuzzi Family Legacy, Grandfather's Advice to Playing the Really Long Game of Compounding
Episode Sponsor: @fiscal_ai is a modern data terminal—link in show notes for a two-week free trial plus 15% off
Find the episode wherever you watch or listen! Thank you, Mike, for a terrific conversation!
If, when you say regulation, you mean the dead and clammy hand of the commissar—the gentleman who has never in his life built a single thing, drafting rules to govern a thing he cannot define, to be enforced by men who cannot read them; if you mean the form in triplicate, the impact assessment upon the impact assessment, the compliance officer who breeds, in the warm dark of the org chart, further compliance officers unto the third and fourth generation; if you mean the moat—the deep cold moat that the giant digs around his own castle and christens, with a perfectly straight face, public safety—the drawbridge he hauls up behind himself the very instant he is across, lest any hungrier and hungrier man should follow; if you mean the precautionary principle, which, had it governed our grandfathers, would have banned the wheel pending further study of the hill, and left us yet shivering and raw in the mouth of the cave, blessing its excellent ventilation; if you mean the European disease—that magnificent open-air museum of a continent, which produces in our time precisely two things in great abundance, and they are regulation, and the eloquent and well-footnoted regret of cultivated men explaining at length why they have produced nothing else; if you mean the license required to think, the permission slip for honest arithmetic, the king’s wax stamp pressed upon the forehead of every new idea before it may draw its first breath; if you mean the agency dispatched, with trumpets, to slay a single dragon, which arrives at the cave, surveys the accommodations, and moves in—and spends the ensuing century laying eggs and devouring the very villagers it was sworn to defend; if you mean the startup that perishes not of the market’s honest verdict but of the filing fee, the genius decamping by the next tide to a freer and warmer shore; if you mean the law that arrives, faithful as the swallows, exactly one whole epoch too late—helmeted, plumed, and magnificently armed—to regulate the stagecoach—then certainly, my friends, I am against it.
But—but, my friends—if, when you say regulation, you mean instead the humble steel guardrail upon the mountain road at midnight, the very thing you curse on the easy days and bless on your knees the one night the fog comes down; if you mean the brakes—for it is the brakes, and not the engine alone, that permit a sane man to drive fast and yet arrive alive—and the buttress, without which no cathedral was ever flung so high, but only in spite of which, but because of which; if you mean the meat inspector, who is the single homely reason a man may eat a sausage in this republic without first composing his last will and testament; if you mean the firebreak cut clean through the forest before the dry season of the burning, the smallpox cordon, the buoy that marks the channel, the rule of the road that lets ten thousand strangers hurtle past one another in the dark at fearful speed and arrive, by its quiet grace, every one of them home; if you mean the honest scale and the true weight, the reason a pound is a pound and a dollar a dollar from Natchez to Nome; if you mean the firm and decent wall between the counterfeit voice and the widow’s bank account, between the deepfaked candidate and the ballot box on the eve of the vote, between the loosed and loveless machine and the schoolyard it neither knows nor pities; if you mean the simple plank of law that says the strong shall not, in the gray dawn, feed the weak quietly into the furnace and sell the rising smoke as progress; if you mean, in the end, the one slender thread of trust without which no citizen will ever dare to use the marvelous thing at all—for where there is no rule there is no trust, and where there is no trust there is no commerce, and a miracle that no man dares to touch is no miracle, but only a handsome and expensive ghost—then certainly I am for it.
This is my stand. I will not retreat from it. I will not compromise one inch of it.
How does SpaceX justify a $2T valuation? You have to believe in space-based data centers.
A data-center satellite is far simpler than a Starlink one. Run the math: roughly 170 Starship launches to put one gigawatt in orbit. Do that in a year and you're launching a Starship every other day.
At ~$20B of cash flow per gigawatt, a few GW up by 2028 starts to make $100B FCF conceivable. A lot on the come. Not physically absurd.
Counterintuitive call: a 5% cash yield is easier to reach at SpaceX ($2T) than at Tesla ($1.4T).
Both need ~$100B of free cash flow to get there. Tesla's path requires operating a national robotaxi fleet plus humanoids at scale. SpaceX's path is launching payload to orbit and getting paid for it, which is the thing it already does.
Launching mass is the lesser challenge than inventing the robots.
New Telltales episode, Data Centers in Space: How SpaceX Justifies $2 Trillion.
Why a launch every other day, not robots, is the more believable path to a 5% cash yield. Plus the oil-to-gas trade, xAI's one-year payback in Memphis, Apple's Siri problem, and a breakthrough in pancreatic cancer.
https://t.co/Fm9JvfYxrq
Who actually runs Iran? Not the government you see on TV.
The Revolutionary Guards are the final arbiters now. They hold the missile and drone inventories, run their own army and navy, and the parliament speaker is one of their own. Hunt puts the odds of an indefinite US-Iran impasse around 50%.
If you are modeling oil through Hormuz, that is who you are actually modeling.
https://t.co/vdbfP5bQkK
This week the AI trade stopped being about the chips and became a question about who pays for them.
Dilute, borrow, or rent. Alphabet sold $85B of stock. Meta floated a raise it doesn't need. Apple chose to pay Google a billion a year instead. Three of the richest companies on earth, three answers to the same invoice.
The cashflow read on the week the AI build stopped paying for itself. New Weekend Update drops tomorrow, plus an updated Telltales Yield and the first issue of The Issue.
https://t.co/r7yhKEYBZE
Two Texas border towns, nearly identical populations. By 2006, McAllen was spending about $15,000 per Medicare enrollee and getting worse outcomes than El Paso.
Here is the part everyone gets wrong: it predates the Affordable Care Act. The cost problem in US healthcare is not partisan. It is structural.
https://t.co/am2r2yyZIp
"Nothing is so permanent as a temporary government program." Milton Friedman.
The Obernolte-Trahan "Great American AI Act" is a 269-page House discussion draft. I didn't read it all myself. I put a swarm of AI agents on it, one analyst per section, plus an adversarial pass that walked back 27 of its own first conclusions before I trusted it. (Using AI to dissect AI regulation, yes, I see it too.)
The detail most readers will note approvingly: the core governance provisions sunset in three years. Light touch. Temporary.
I wouldn't bet on it. The sunset is what makes the bill passable, not what makes it temporary. Underwrite the permanent version: a regulated perimeter drawn around the largest labs, a licensed audit industry, a new standards body at Commerce. Here is what that version does.
1. The preemption is not what the headline will say. Everyone will call this "federal preemption of state AI law." Read the text: it preempts only state laws that specifically regulate model development, training, evals, pre-deployment gating. It expressly preserves common-law tort, consumer protection, and all post-deployment product liability for the states (savings clauses in 121(c)). It clears underbrush on the input side. It does not insulate a deployed model's behavior from state-court liability. Don't trade the "this de-risks AI product liability" headline. It doesn't.
2. The industry "shield" is narrower than it reads. The anti-regulatory-use and FOIA protection (102(d)) covers only data a company voluntarily shares in cooperation. It does not immunize the third-party audit pipeline (Section 112), which is exactly where the real enforcement exposure lives.
3. The public-compute build gets misread in both directions. The National AI Research Resource is neither a single-cloud win nor a frontier-lab lockout. "Multi-vendor" is an NSF design preference, not a binding mandate, and the startup-only access cap (under 7 years old, under 500 employees) is a default the agency can widen, university-affiliated researchers included. Demand-side color for hyperscalers, GPU vendors, and the cooling supply chain, not a competitive constraint.
4. The open-source access mandate carries a live overhang. Large frontier developers can be compelled to give designated open-source security maintainers free model access, gated case-by-case by CISA. The catch: the "reasonable controls" meant to protect the developer's IP are set by CISA's rulemaking, not by the developer. A real, rulemaking-contingent IP risk, not a footnote.
Net: less a safety bill or a preemption bill than scaffolding for a permanent, incumbent-favorable, disclosure-and-audit AI regime, sold with a three-year sunset to get it through. The five labs big enough to absorb it get a moat. A new licensed audit industry gets born. And if Friedman was right, the "temporary" part is the part that lasts.
Curious how others who've read it land, especially on whether the preemption survives markup and at what scope.
Same knee replacement, same region: $23,000 at one hospital, $103,000 at another.
Montana's state health plan set reference-based pricing at 2x Medicare and told hospitals to take it or be excluded. They saved a fortune, cut zero benefits, and every hospital eventually fell in line.
Then the administrator retired, the lobbyists moved in, and it reversed. The system follows the incentive, not the hero.
https://t.co/H3mIrW6VI3
New Telltales episode is up.
The ACA built the greatest insurance business ever conceived: a guaranteed 15-20% margin on cost. When you earn 15% of the bill, you want the bill to be bigger. That one incentive explains most of what is broken in US healthcare.
Hunt, Jason and I work through it, plus oil through a contested Strait of Hormuz and the deficit math that could eventually force Medicare for All.
https://t.co/XkXJuOo6K2
NVIDIA just set a new all-time free cash flow record. ~$163B on a run-rate basis. That eclipses the old Exxon peak.
For scale: Apple is at ~$120B. Both at multi-trillion market caps.
NVIDIA at $5.6T. Apple at $4.5T. Two FCF machines that didn't exist at this scale a decade ago.
One of the new, buzzy jobs in Silicon Valley is the AI Forward Deployed Engineer (FDE), an engineer who is embedded within a client organization to help customize solutions, such as building and tuning agentic workflows that suit the client’s particular needs. I’ve heard from people who are wondering anew about the FDE career path since OpenAI and Anthropic started building new teams to place FDEs within client organizations.
The rise of FDEs for AI workloads is one way AI is creating new jobs (and why the jobpolcalypse narrative of upcoming job market collapse is false -- there will be many AI and non-AI jobs). However, I believe there will be far more AI Engineer jobs than FDEs, as I explain below.
The FDE role was pioneered about two decades ago by Palantir, which sent engineers to government locations to work on secure, air-gapped networks. In addition to having good technical skills, FDEs need communication skills and sometimes business skills. For example, they may need to speak with clients to understand their needs, formulate a strategy to prioritize projects, explain complex technology, and respectfully push back if a client asks for something unrealistic. They’re enjoying a resurgence because of the amount of work involved in taking an off-the-shelf LLM and building it into a custom agentic workflow that fits particular business needs.
However, I believe the number of AI Engineer jobs will be far larger. A company might accept a few FDEs to be embedded within its organization. But most companies will want far more of their own employees working on their projects. While my organizations do hire FDEs, we hire far more AI Engineers! Also, a common client concern is that it is hard to find vendor-neutral FDEs — they are, after all, there to deeply integrate a particular vendor’s product into a company. In this moment when it’s hard to predict which AI service will be the best one in a year’s time, optionality (the ability to pick whatever vendor turns out to fit best in the future) is very valuable. In contrast, letting FDEs tightly bind a company’s processes significantly reduces optionality.
Right now, I see surging demand for AI Engineers who can build software applications using AI software components (like LLM prompting, agentic frameworks, evals, etc.) and effectively use AI coding agents (like Claude Code, Codex, Antigravity CLI, and OpenCode). As the AI Engineer role matures, I expect it to fragment into more specialized roles, like the generic Software Engineer role from decades ago fragmented into frontend, backend, mobile, data engineering, devops, and so on.
What will be the future, specialized AI engineering roles? I don’t know. Perhaps there will be AI FDEs, LLMOps Engineers, Evals Engineers, AI Data Engineers, Harness Engineers, and other roles we don’t have names for yet. But for now, I see a lot of AI engineers who are generalists create a lot of value. Skilled AI Engineers are in very high demand! As our field continues to mature over the coming decade, I look forward to new specializations within AI Engineering that create even more job opportunities.
[Original text: The Batch newsletter]
NASA: ~$19,000 per kilogram to orbit.
SpaceX Falcon: ~$850/kg.
SpaceX Starship: ~$100/kg.
SpaceX is spending $4B on Starship R&D this year, up from $3B. Customer-launch margins are 65-75% and rising as cost per kilogram falls.
This is why launch — not Starlink — is the crown jewel.
Bogumil Baranowski (@bogumil_nyc) had me on Talking Billions — an hour he titled "Sailboats, Telltales, Roads & Motorcycles."
We got into how Top Mark Capital actually started: I was prepping a boat for an offshore race and noticed a group in a back room reading 10-Ks. The introduction out of that room became the firm and the podcast.
The lesson wasn't "follow your passion." It was narrower: opportunities compound through people, and people compound by doing new things together.
https://t.co/G3dJDqwfJo
Apple in 16-17 had not quite the same, but plenty of institutional coverage and yet was left for dead. Google, similarly just months ago. Plenty of opportunities for the stock picker in today’s markets regardless of mkt cap if you have the insight. Especially as passive indexing intensifies.