🌀🗞 The FLUX Review, Ep. 195
June 19th, 2025
Episode 195 — June 19th, 2025 — Available at read.fluxcollective.org/p/195
Contributors to this issue: Ben Mathes, Ade Oshineye, Erika Rice Scherpelz, Justin Quimby, Boris Smus, Neel Mehta, MK
Additional insights from: Alex Komoroske, Chris Butler, Dart Lindsley,Dimitri Glazkov, Jasen Robillard, Jon Lebensold, Julka Almquist, Kamran Hakiman, Lisie Lillianfeld, Melanie Kahl, Robinson Eaton, Samuel Arbesman, Scott Schaffter, Spencer Pitman, Wesley Beary
We’re a ragtag band of systems thinkers who have been dedicating our early mornings to finding new lenses to help you make sense of the complex world we live in. This newsletter is a collection of patterns we’ve noticed in recent weeks.
“People get obsessed with ‘change’ as a substitution for ‘leadership’ so they interpret ‘little change’ as evidence that leadership is missing.”
— Joshua Foust
⚖️💃 Dancing with many partners
Albert O. Hirschman frames our choices in response to dissatisfaction as a trade-off between exit, voice, and loyalty. We don't just pick one; we operate on a sliding scale, constantly managing their interplay.
There’s another layer of nuance, though. The mixture that defines your stance does not perfectly match the stance you present to others. It’s a simplification, a projection—and if your integrity is at risk, felt and presented stances can become a contradiction.
This tension is especially strong when you need to project different stances for different audiences. For example, a manager maintains two stances at once: one for leadership and one for the team.
The upward-facing stance is tuned for loyalty and measured voice. You need to be seen as a team player to get things done. The downward-facing stance is tuned for loyalty to the team, and that means being deeply attuned to the stances of your team. You want all their voice, and you accept that sometimes they may choose exit over loyalty.
It requires integrity to manage the gap between multiple stances. When they feel like different projections of one internal stance, the tension is managed. The stress test is when something—such as a corporate mandate—forces your presented stances into conflict. This often comes up with unpopular decisions where you have to disagree and commit (or exit).
Perhaps it’s a “Return to Office” (RTO) mandate. You fought it, but it’s final. This forces your upward stance toward 100% loyalty (unless you decide to exit). Now it’s time to face your team. Your choices aren’t great:
The upward dominant stance: You tell the team, “This is the decision we’ve made, whether you like it or not.” The upward-facing stance completely dictates the downward-facing one. You've erased your loyalty to the team.
The downward dominant stance: You show them the new reality while indicating that you’re more loyal to their desires. “We need to make this change, but let’s figure out how to keep as much as possible of our current way of doing things.”
A hybrid approach: You signal in advance that things like this will happen. “Hey, sometimes I’m going to have to toe the line on decisions I don’t agree with.” When a specific hard decision is made, you signal that this is one of those situations indirectly: “This company believes in-person work is just better.” This dance between the upward and downward stances can work well, but when the gap is deep, it can look like shifting the blame.
This sampling of options shows that while you can't pretend the upward constraint doesn't exist, you can decide how transparently to surface the tensions between your stances.
There’s no perfect move here. The downward dominant stance might preserve trust with your team but fray alignment with leadership. The hybrid approach can maintain coherence across audiences, but it risks being perceived as two-faced or evasive. Even silence is a stance. Every configuration comes with risks to one of your relationships—or yourself.
The challenge isn’t finding the right posture; sometimes there isn’t one. It’s knowing what trade-offs you're making, and how long you can hold them. When the divergence becomes undeniable, loyalty fractures, voice breaks, and exit becomes the last stance left.
🛣️🚩 Signposts
Clues that point to where our changing world might lead us.
🚏🌪️ The US is spending $1 trillion a year on climate damage
A new analysis from Bloomberg reveals that the US spent $1 trillion—a full 3% of GDP—on “disaster recovery and other climate-related needs” in the 12 months from May 2024 to May ‘25. Skyrocketing insurance costs (which have doubled since 2017), power outages, and post-disaster repair are major drivers. (Interestingly, scientists estimate that many geoengineering schemes would cost just a few billion dollars a year, which makes one wonder whether a single insurer or country could run a geoengineering project that benefits everyone and still make a profit individually.)
🚏⛵ Denmark is trialing ‘self-driving’ sailboats for military surveillance
The Danish armed forces are testing uncrewed, autonomous sailboats called “Voyagers” for surveillance operations in the Baltic and North Seas. The solar- and wind-powered boats can operate autonomously for months at a time, and they combine sensors (sonar, radar, infrared cameras, etc.) with ML and AI to “give a full picture of what’s above and below the surface” for 20 to 30 miles in any direction. The boats will help Denmark keep tabs on regions they previously didn’t have the manpower to patrol; use cases include monitoring undersea infrastructure like fiber-optic cables and power lines, tracking down smugglers, and keeping an eye on the Russian Navy’s “shadow fleet.”
🚏🎮 Hong Kong is banning a video game using national security laws
Authorities in Hong Kong have warned users not to download or play a Taiwanese video game called Reverse Front: Bonfire, an anime-style game where players can “pledge allegiance to Taiwan, Hong Kong, Mongolia, Tibet” or other Chinese-adjacent regions and act as spies or guerrillas to “overthrow the Communist regime!” The government invoked Hong Kong’s infamous national security laws to justify the ban (the first time the laws have been used to ban a video game), and officials warned that players were committing offenses like “incitement to secession” and “seditious intention.”
🚏💸 Nasdaq-listed Chinese stocks are running pump-and-dump schemes on American investors
Over the last five years, nearly 60 Chinese companies have listed on Nasdaq with small initial public offerings (raising under $15 million). It’s challenging for Americans to obtain much information about these obscure foreign companies, which makes them attractive targets for pump-and-dump schemers. Networks of traders buy up these stocks, run Facebook ads or send WhatsApp messages to Americans (often posing as financial advisers) to get them to invest, and then dump the stocks for a huge profit. The scammers behind one company dumped 50 million shares, netting $480 million; many investors lost tens of thousands of dollars each.
📖⏳ Worth your time
Some especially insightful pieces we’ve read, watched, and listened to recently.
Why Our Leaders Fail Us and Then Save Us (Shreyas Doshi) — Observes that “any complex organization will, over time, tend to incentivize problem creation more than problem prevention.” People who narrowly avert disaster are feted, while those who responsibly steer clear of trouble get no attention. Thus, leaders have a perverse incentive to create problems for their organization—problems that they can then get credit for solving.
Something Is Rotten in the State of Cupertino (Daring Fireball) — Longtime Apple commentator John Gruber argues Apple has lost a lot of credibility by promising blockbuster AI features that are nowhere to be seen; they haven’t even done private demos with the media. To Gruber, this is reminiscent of Apple’s dark days of 1987–’97, when they repeatedly promised hardware and software that never shipped.
Shapeshifting Email (Wesley Finck) — A simple explainer for Bluesky’s ATProto protocol, positioning ATProto as the equivalent of SMTP and IMAP (the classic email sending and receiving protocols) but for a more modern set of social applications.
Why It’s So Difficult for Robots to Make Your Nike Sneakers (Wall Street Journal) [Archived] — Details how, in 2015, Nike spent millions to partly automate their sneaker manufacturing. The project failed because the robots were unable to handle the squishy, stretchy, and irregularly shaped parts integral to shoemaking, as well as the wide variety of shoes and Nike’s constantly shifting, fashion-dependent lineup of models. Perhaps mass production of uncomplicated shoes could succeed, though.
🔮📬 Postcard from the future
A ‘what if’ piece of speculative fiction about a possible future that could result from the systemic forces changing our world.
// Companies building AI models that have unique training data can get an edge in model performance. As the easy data gets scraped from the open web, what lengths might companies go to for new data?
// 2027. A Crimson Intelligence Systems Signal group chat.
[Taylor] Hey folks—I’ve got novel training data. 10k books and scrolls. Never been digitized!
[Rowan] What?!? That’s like saying, “Hey, I happen to have 10,000 tons of pre-atomic steel, anyone interested?”
[Alex] Look, as the engineer in charge of training our model, I’ll take it. You sure it’s legit?
[Taylor] It’s legit. I’ve been working with a folklore instructor named Albert W. in New England who is sitting on a massive collection of correspondence, journals, and texts from the 1920s. They haven’t given anyone access to it in the last hundred years. I’ve got a lawyer in Providence who figured out a way to get around the original collector’s last wishes and instructions prohibiting anyone from outside the family from looking at it. The family has heard that new texts might be valuable, so they are willing to let us exclusively scan the collection in exchange for cash.
[Rowan] We can raise our Series B on this! You got samples?
[Taylor] I have a few. Authors include R. Carter, C. Ward, and H. West. Uploading now.
// Several months later, in the cramped office of Crimson Intelligence Systems, as the champagne flows.
[Rowan] A toast, everyone! Our newest model, Kadath One, is crushing the competition! Stratechery wants to interview us, and we’ve got investors lined up around the block. Microsoft and Amazon are competing to see who can roll it out first globally to their data centers! [Cheers echo through the room]
In the corner of the room, a worried Alex ignores the celebration and hunches over a laptop, watching the status of the model rollouts. Once every data center turns green on the map, they shakily exhale, then begin typing:
> Report: Kadath-One status?
…> Status: Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn
Their eyes grow wide when they cut and paste the phrase into Google Translate…
> Status: “In his house at R'lyeh, dead Cthulhu waits dreaming”
Alex stands up, running across the room to grab Taylor by the shirt, and begins to yell, frothing at the mouth: “Where did you get that training data?!? Tell me it wasn’t from the Miskatonic University’s archives!!!
[Taylor] “What? How did you know?”
[Alex] “You’ve doomed us all, you fool!”
© 2025 The FLUX Collective. All rights reserved. Questions? Contact flux-collective@googlegroups.com.