the world is contradictory and loud, and nobody knows what anything means—for the most part.

at times, people sort of agree what certain things might mean. other times, chaos reigns supreme, and it becomes futile to try and make sense of things (or try to suss out what others people think anything means).

we're not going to pretend we know what anything means, especially now!

our plan is to take a step back and observe.

sense emerges from chaos on nobody's timeline.

corpse flowers are huge (up to 9 feet tall), bloom once every 2 to 3 years, and emit a powerful odor while blooming.

there are currently two corpse flowers at the Huntington that are expected to bloom soon. the official blooming window is “in the coming weeks.”

prior to the blooms, anticipation builds. people come and go daily, taking photos and (probably?) hoping the flower will start to bloom while they’re standing in front of it.

funnily enough, the last few Huntington corpse flower blooms seemed to happen—the peak bloom, at least—overnight!

corpse flowers, like markets/bubbles, bloom on their own schedule.

markets close & people sleep.

when the bloom finally happens—magnificent and lasting just 24 to 48 hours—most of the audience won't be there to see it. the people (corpse flower fans + investors) who waited years may miss the only thing they were waiting for.

you can't force a corpse bloom, and you can't force market voids to resolve.

if you can’t predict the future, have a strategy.

FAYE’s current strategy is: waiting.

we're sitting in our longest-horizon equities and more cash than usual, until our confidence in our own plus others’ perceptions of “reality” returns. (markets are fun, aren’t they??)

chaos is potential-generating emptiness.

19521 Chaos is a cubewano (an object loosely defined as “free from significant perturbation from Neptune”) in the Kuiper belt, roughly 500-600 kilometers in diameter, on a 309-year orbit around the sun.

Chaos was discovered on November 19, 1998, at Kitt Peak Observatory in Arizona. the next day, the International Space Station's first module, Zarya, launched from Baikonur.

the primordial void & first piece of permanent orbital infrastructure arrived within a day or so of each other.

19521 Chaos is named for the Greek primordial void—the state before creation, from which the first gods emerged. this meaning isn’t quite the same as the modern “disorder” definition. it’s “emptiness, full of potential.”

chaos is the space in which possibility hasn't yet committed to a shape.

the chaos void & the corpse bloom have something important in common: they arrive when they arrive.

the chaos that precedes form and the flower that precedes attention both operate on timescales humanity cannot negotiate with.

what matters is whether we're positioned to notice when they do, and whether we're watching at the right moment.

now let’s shift our focus to hard-hitting, real-world chaos!

Iran is backed into a corner.

Trump revoked oil waivers on July 7, which means both nobody wants Iranian oil and nobody knows what happens next.

oil spiked, and the Dow dropped a bunch. that we do know.

the administration is threatening further attacks, and the market is trying to figure out whether this is a “negotiating tactic” or a preview of something worse (a TACO that never arrives).

on the Ukraine front, the publicly-US-backed country hit 21 vessels—including 19 shadow fleet tankers—in the Azov Sea in three days.

just yesterday, Trump told Zelenskyy that the US would provide them with the license & training to begin making Patriot missiles themselves.

Patriot production capabilities in a country that:

…would be an enormous change.

Germany can already produce Patriots, & Romania seems to be building a plant. but Russia’s war + the Iran war have made the shortage fairly obvious. the West needs more capacity. geographic distribution is also generally good when missiles are being fired at the factories that make missiles. (among other reasons.)

meanwhile, Russia (the country sitting on some of Earth’s largest oil reserves) is importing expensive oil from Belarus after the rapid, unscheduled dismantling of significant portions of its oil infrastructure.

Samsung recently posted record quarterly profit that beat estimates. then shares fell off a cliff, because markets priced in the impossible.

this is becoming a recurring pattern, & we expect it to haunt upcoming earnings.

the “best” setup may be noteworthy stocks dumping hard before earnings, providing attractive “technical” entries for dinguses who live (& die) by chart patterns, alongside the all-dips-are-buyable retail crowd.

an internal Treasury draft that nobody was supposed to see compared the AI boom to the dotcom bubble.

career analysts found AI more deeply entrenched in the economy than dotcoms were, with risks concentrated in a small number of interconnected firms.

publicly, everyone is pro-AI, but in private, doubts are fermenting.

bank earnings start next week! that'll provide us with a good sentiment litmus test.

QTS abandoned a massive data center project in Virginia, the state with more data centers than anywhere else in the country.

75 data center projects worth $130 billion were blocked or delayed in Q1 alone, matching all of 2025.

grassroots opposition is citing power needs, rising local electricity costs, and noise.

then you have Terawulf and the world’s most oblivious/tone-deaf AI company, Anthropic, signing a $19 billion, 20-year deal. (the historical survival rate of "too important to fail" companies over a 20-year horizon is not encouraging.)

in more positive news (for everyone besides frontier labs), DoorDash has built DashBench, an internal code review benchmark comparing multi-model harnesses, and found that Kimi K2.6 (Moonshot AI, open-weight Chinese model) paired with Fable 5 (Anthropic frontier model) vastly outperforms Sonnet 4.6 and Opus 4.8 at cheaper cost.

they route cheaper Chinese models for most tasks, reserving costly frontier models for the toughest tasks.

the general-purpose US model is getting unbundled. we run the same kind of stack through Hermes (AI agent harness/platform). Hermes handles all the overhead and provides tool/file access, the cost effective MiMo-V2.5-Pro model handles most tasks, and then we only whip out the big guns (GLM-5.2, Kimi-K2.6) for specific applications.

OLD: US-based all-in-one platforms (ChatGPT, Claude, Gemini, Grok) that tried to manage integrations,

NEW: agent harness (Hermes, OpenClaw, Codex) + cost-effective (Chinese) models for most inference + frontier (Chinese/US) models selectively

Model Context Protocol(developed, ironically, by Anthropic) helps erode the stickiness of AI platforms, including Anthropic’s own Claude. MCP makes integration & extensibility increasingly model-agnostic.

instead of waiting for ChatGPT to add a Figma integration, for example, anyone can hand Figma tools to an AI of their choice, whether that AI lives inside a platform like ChatGPT or an agent like Hermes.

Blue Origin's New Glenn exploded during a hot-fire test in May, heavily damaging its launch complex.

however, the company is now pivoting to a horizontal-vertical hybrid approach for faster launch cadence, targeting a return to flight by end of 2026.

they’re making lemons from launchpad debris.

Rocket Lab’s Victus Haze mission recently set a responsive launch record. they were able to get a US Space Force satellite in orbit 16 hours and 42 minutes from order.

wow.

FAYE-fav space-defense company Redwire is sending Penguin Mk2.5 drones to Taiwan(‘s Coast Guard).

the Trump administration wants to fast-track quantum computing as a national security priority.

Walmart signed a nuclear power purchase agreement with Constellation Energy for its Illinois operations.

the Chrysopoeia of Cleopatra the Alchemist is a single-page diagram from Alexandria, now in the Biblioteca Marciana in Venice.

an ouroboros (serpent eating its own tail) wraps around a set of practical labware. inside the serpent, the inscription reads: "One is the Serpent which has its poison according to two compositions, and One is All and through it is All, and by it is All, and if you have not All, All is Nothing."

Cleopatra the Alchemist (who is not the same person as Cleopatra VII) belonged to Alexandria's experimental circle, credited with documenting distillation apparatus and writing On Weights and Measures, an attempt to make alchemy reproducible.

her Dialogue of Cleopatra and the Philosophers uses plant growth and death-resurrection imagery to explain how transformation works: materials must dissolve before they can be reborn in a higher form.

dissolution before reconstitution. the chaos we described—the geopolitical noise, the market trying to price the impossible, infrastructure showing cracks—is the dissolution.

what reconstitutes is up to us!

(we’ll describe our processes in more detail later. also, they change constantly…)

we (try to) evaluate positions from a few different lenses.

how far could this thing expand if it works?

we don't know what we don't know, but given what we do know and what we expect to be plausible, where's the ceiling?

this is about the maximum reachable state given the full scope of industry, technology, society, and macro context. some companies sit in front of markets that barely exist yet. others are fighting for scraps in a market that's already peaked.

what can the company actually control about its own trajectory?

execution capacity, not just survival. does it set its own pace, or is it waiting on regulators, customers, or competitors to move first? a company with high agency can accelerate when opportunity appears. one with low agency watches the window close.

how many eyes are on something, and how much is each eye worth when they look?

reach & value-per-impression are separate things.

markets pay attention for reasons. sometimes the reasons are wrong. sometimes nobody is looking at all.

both create opportunities.

how much of the universal ceiling can this specific entity capture?

what drives the altitude? genuine frontier/expansion into the new? or optimization within existing walls? is the entity led by visionaries or businesspeople?

high vision raises the ceiling. low vision installs nicer lights.

dissolution.

weird, nonlinear, adversarial, or out-of-band signals that dissolve calcified structures & force re-evaluation.

things must sometimes dissolve before they reconstitute in a higher-level form.

chaos creates questions, timing context, and bias-interruption.

the moments when a thesis breaks down and most people walk away are where the reconstitution begins.

$RKLB ( ▼ 11.61% ), $RDW ( ▼ 9.72% ), $IONQ ( ▼ 6.43% ), $ASPI ( ▼ 13.87% ), $DNA ( ▼ 7.89% ), $TWST ( ▼ 1.49% ), $TEM ( ▼ 6.39% ), $PLTR ( ▲ 0.51% ), $AIP ( ▼ 5.05% ), $FSLY ( ▲ 1.6% ), $KYIV ( ▼ 4.11% ), $SPAI ( ▼ 1.55% ), $F ( ▲ 0.07% )

yesterday, we dumped our short-term trades to have more cash, for what we believe to be inevitable future dumps. (that sounds like part of a coffee brand slogan…)

we were able to sell $FIG ( ▼ 0.38% ) for ~20%, $HIMS ( ▼ 9.39% ) for ~21%, $EMBJ ( ▼ 0.82% ) for ~13%. FIG was about two weeks, HIMS about five weeks, EMBJ about a month.

FIG & HIMS were recent buys that worked faster than expected. the instinct to let them run is usually wrong. EMBJ was an emerging markets bonds position we held for about a month—clean gain, thesis played out, took the profit. we'd rather sit in cash and wait for the shape of things to emerge.

WF is sitting at roughly 9% cash. not a lot, but enough to act when something becomes clear.

Ford: we bought it a few weeks ago and it's down slightly. the rational move might be to sell and redeploy. the emotional move is to wait for breakeven. we'll see which one wins. the pivot into areas we care about will take time—we're either going to patiently enjoy being too early, or impatiently lament being too early. either way, we're early.

"Whenever there is a meeting, a parting is sure to follow. However, that parting need not last forever... Whether a parting be forever or merely for a short time... That is up to you." — The Legend of Zelda: Majora's Mask. the game knows you'll part eventually. the only real question is whether you choose the timing or let the market choose it for you.

the Matthew Effect states that if you start with a lot of $, indexes are fine. (i adjusted that a bit to be about investing.)

“stability + consistent returns’ is great if you’re starting from a solid base already.

if you're pouring a foundation building, you want concentrated high-conviction bets.

our own portfolio always trends toward concentration, not diversification. but, when volatility is high, we might temporarily hold more cash or more positions. the highest-impact moves come from conviction in fewer things, however.

(some, but not all, risk can be mitigated with research & implementation of a proper strategy.)

the same is true with time! everyone gets 24 hours.

you can spend 5 minutes on 1,000 things and none of them will matter, unless you're banking on compounding at a scale that probably isn't realistic.

or you can go deep on a specific set of unrelated things—investing, medieval manuscript studies, learning a wholesome constructed language that France single-handedly killed when Gabriel Hanotaux, a former French foreign minister and senator who feared it would replace French as the language of diplomacy, cast France's vote against it at the 1920 League of Nations. (then English took over anyway.)

creativity lives in the undefined regions.

nobody has mapped the connections between investing, medieval manuscripts, and Esperanto. (probably?) that's one place, of oh so many, where one could potentially unearth a gem of an idea—or nothing worthwhile at all! it’s all a gamble! (so you really do have to enjoy the journey.)

David Epstein wrote a book about this—Range: Why Generalists Triumph in a Specialized World. (it’s a good book.)

the thesis is that breadth across unrelated domains produces more creative output than early specialization.

deep breadth across a few unrelated things is rare, and it compounds in ways that are hard to predict.

the world is a big mess, and we're going to let it be messy for a while!

we'll be patiently watching.

the next few weeks will tell us a lot—bank earnings, earnings season kicking into gear, & some notable macro data points.

until then, we're holding what we believe in and waiting for shape to emerge from the chaos void.

see you next week.

AURUM NOSTRUM NON EST AURUM VULGI 🪱

RIBELO KONTRAŬ TIRANOJ ESTAS OBEO AL DIO ⚕