🌀🗞 The FLUX Review, Ep. 167
October 24th, 2024
Episode 167 — October 24th, 2024 — Available at read.fluxcollective.org/p/167
Contributors to this issue: Neel Mehta, Boris Smus, Dimitri Glazkov, Erika Rice Scherpelz, Ade Oshineye, Dart Lindsley, MK
Additional insights from: Ben Mathes, Justin Quimby, Alex Komoroske, Robinson Eaton, Spencer Pitman, Julka Almquist, Scott Schaffter, Lisie Lillianfeld, Samuel Arbesman, Jon Lebensold, Melanie Kahl, Kamran Hakiman, Chris Butler, 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.
“Rest is not idleness, and to lie sometimes on the grass under trees on a summer's day, listening to the murmur of the water, or watching the clouds float across the sky, is by no means a waste of time.”
― John Lubbock
🌱🌳 Mighty oaks from little acorns grow
The world is complex, and communicating complexity is challenging. The mysterious language and abstract concepts of systems thinking are just as likely to confuse an audience as they are to empower it. Terms like “feedback loops,” “slime molds,” and “kayfabe” seem disconnected from their familiar usage. Some people just don’t care about that nuance. Others, while recognizing the importance of nuance, need solutions now. This is where what our group calls “defluxing” comes in. Defluxing breaks down complex systems concepts into their simplest, most relatable forms, using accessible language and familiar narratives.
Defluxing can feel quite dispiriting. Yet, sometimes, change requires broader buy-in. And with that, there is a need to make systems thinking not just palatable but actionable. The butterfly effect? Talk about how a small decision ripples through an organization. Emergence? Show how different team members' inputs lead to unexpected solutions. You’re not diluting the idea — you’re making it accessible.
Defluxing is a tool for a systems approach to communication. Ideas live in ecosystems and need the right conditions to take root. Removing a bit of nuance increases the chances that seeds will grow in places that resist a fully formed idea.
Defluxing has risks: ironing out nuances often leads to oversimplification — a plastic tree instead of a sapling. While conveying an idea in a simplified form is usually better than not at all, there is a risk of falling into cargo culting, where solutions are applied without understanding their context. Systems thinking resists one-size-fits-all answers.
How do we deflux in practice? Ideas, like memes and viruses, need to evolve for broader adoption. Defluxing follows a familiar evolutionary pattern: iterate, adapt, and remix.
First, iterate by experimenting with different ways to communicate an idea. Start with an example or a metaphor. How about a personal story? Test what resonates and refine the message based on feedback. Adapt by looking at the context. Who is the audience? What problem are they trying to solve? What would make the message relevant to them? Remix by combining elements from different sources, disciplines, or perspectives to make ideas more engaging. You can pull from theoretical physics... or a Taylor Swift song.
At its best, defluxing isn’t about watering down. It’s about finding the heart of an idea. By iterating, adapting, and remixing, we find ways to invite others into the conversation. It is an exercise in gardening — engaging the world one practical, intuitive insight at a time.
Applied thoughtfully, defluxing nurtures systems thinking by creating the right conditions for growth. Each small, carefully planted insight can take root and flourish, leading to a more significant paradigm shift... or at least a good idea or two being put into practice. And occasionally, you'll forge a new connection with someone else who “gets it.” By cultivating these ideas and relationships, systems thinking can grow quietly, becoming an integral part of understanding and shaping our everyday world.
🛣️🚩 Signposts
Clues that point to where our changing world might lead us.
🚏🗳️ More states are setting early voting records
Last week, we discussed how the state of Georgia had smashed its record for the first day of in-person early voting. Other US states have since joined the party: almost two million people in North Carolina cast a ballot in the first six days of early voting, breaking records from the past two presidential elections, and early voting on Wisconsin’s first day blew past the comparable figure from 2020. Nationwide, over 30 million Americans have voted early.
🚏🐐 A VC-backed AI bot pumped a meme cryptocurrency and became a ‘millionaire’
An LLM-powered bot called Truth Terminal started posting and responding to people on Twitter/X, and it got a $50,000 grant when it pitched itself to venture capitalist Marc Andreessen. Later, fans of the crypto ‘memecoin’ Goatseus Maximus asked the bot for an endorsement and sent the $GOAT token to the bot’s crypto wallet, and the bot started making pro-GOAT posts in return. The coin’s price soared (its market cap is about $700 million at the time of writing), making the bot a millionaire, at least according to the market price of the crypto-coins in its wallet. (Note that the human creator of the bot controls the wallet and approves LLM-generated posts before they go live, which has raised questions about how autonomous the bot really is.)
🚏🌧️ A tech company says its “cloud exit” will save it $10 million
37Signals, the tech company behind the productivity tool Basecamp, announced in 2022 that it would be pulling itself off major cloud hosting providers, instead buying its own servers and storage flash arrays. The company’s outspoken co-founder, David Heinemeier Hansson (also known as “DHH”), shared that the company had bought $800,000 in gear but expects to save about $10 million in cloud costs over five years. He added the caveat that cloud vendors are still essential for new companies or those facing “enormous fluctuations in load.”
🚏🛐 Online-only churches are gaining popularity in Kenya
Many houses of worship were forced to go online during the pandemic, and today, some Kenyan preachers have taken that to its logical conclusion by launching online-only churches that feature nightly sermons livestreamed over platforms like Facebook and TikTok. Congregants can send donations through Kenya’s popular mobile payment service, M-Pesa. Buoying the trend is the fact that internet users in Kenya spend almost four hours a day on social media, more than any other country, and that Kenya has a large diaspora population in places like Saudi Arabia and the United States.
📖⏳ Worth your time
Some especially insightful pieces we’ve read, watched, and listened to recently.
The 10 Most Important Rules of Storm Chasing (June First) — Through the lens of chasing dangerous tornadoes, this video essay explores how we should make decisions in fast-moving environments with limited information. Maintain your situational awareness (and be cognizant of when you don’t have it); don’t fixate on signals that can quickly get stale (like weather reports on your phone); and avoid “task saturation” so you don’t get mentally overwhelmed.
Beyond Lean and Agile (Marty Cagan) — Simple yet timeless advice for startups building MVPs: tackle risks up-front, design products collaboratively, and focus on solving real problems.
Market Prices Are Not Probabilities (Quantian) — A somewhat technical deep-dive into why electoral betting markets aren’t good predictors of electoral outcomes. Fundamentally, the “price” of a given candidate winning just tells you what people want to pay for that contract, warped by limits to arbitrage, animal spirits, outsized payoffs for long-shot bets, etc. That market-clearing price doesn’t necessarily have any connection to the “fair” value (a given candidate’s actual chances of winning). In short, prices aren’t probabilities, at least when betting on things that haven’t happened yet.
America Needs More Weird Homes [Archived] (Business Insider) — “YIMBY” writer Ned Resnikoff argues that the bland uniformity of American housing (including the dreaded 5-over-1 apartments) is due to the US’s overly rigid zoning codes. By streamlining zoning policies and legalizing more varied types of construction, we can let housing development adapt to the needs of local markets — which will give people more housing options and make our cities more architecturally interesting.
🔍📡 Lens of the week
Introducing new ways to see the world and new tools to add to your mental arsenal.
This week’s lens: antennas and probes.
We rely on tools to explore and understand our increasingly strange world. The metaphor of antennas and probes suggests two complementary tools for exploration.
Both antennas and probes are sensing devices. Antennas stand tall and proud, letting the entire cacophony of the signals wash over them. Probes, by contrast, embed themselves deeply into their subject, picking up specific signals right where the action is.
Antennas, with their breadth and scale, are excellent for big-picture analysis. When we want to understand trends, they give us an “everything bagel” view of the environment. Researchers might spend their days scanning the latest papers and press on generative AI. They can use this broad sense to create slide decks with helpful information about what might happen in that space.
But the key word here is might. Antennas are susceptible to interference. Their insights likely come with a good dollop of hype alongside their substance. Separating the wheat from the chaff takes a lot of careful fact-checking and additional research. This uncertainty means that antennas tend to have a slow feedback loop. Transmission, translation, and denoising take time. Antennas give signals that trail behind reality, especially in fast-moving fields.
Probes focus right on the source, capturing the original signal right as it happens. For instance, instead of reading about a cool AI project in the news, a probe approach involves playing with the project, looking at the code, and contributing. Probes directly engage with the real deal. The downside of a probe is that we might be looking in the wrong place. Misplaced probes generate a deep, accurate, but ultimately irrelevant signal.
With a well-placed probe, we get ahead of the curve. Instead of reading a famous blogger’s rant about some new LLM model weeks after it comes out — the antenna view — probes let us go to the Discord server to see the real-time struggles of early adopters. The challenge with probes is discovery. We need access to that Discord server. We need to know where the hot projects are on GitHub. This takes expertise… and some luck.
A well-designed sensing system mixes antennas and probes. We can intentionally choose which tools to use, benefiting from each other's strengths to help us understand our weird world.
© 2024 The FLUX Collective. All rights reserved. Questions? Contact flux-collective@googlegroups.com.