🌀🗞 The FLUX Review, Ep. 97
Episode 97 — April 27th, 2023 — Available at read.fluxcollective.org/p/97
Contributors to this issue: Scott Schaffter, Gordon Brander, Ben Mathes, Erika Rice Scherpelz, Neel Mehta, Boris Smus, Dimitri Glazkov
Additional insights from: Ade Oshineye a.r. Routh, Stefano Mazzocchi, Justin Quimby, Alex Komoroske, Robinson Eaton, Spencer Pitman, Julka Almquist, Lisie Lillianfeld, Samuel Arbesman, Dart Lindsley, Jon Lebensold
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.
“The thing about staying above the API is that the API level keeps going up.”
📈 ⚖️ Mess up anything with this one weird trick
In almost any domain, if we try to maximize any one dimension, we tend to generate unbalanced extremes. Social media focuses on maximizing engagement and thus incentivizes increasingly extreme content. Fandoms and topical communities — Marvel, weight lifting, hair care, whatever — go so deep on their topic that it’s often hard for the uninitiated to catch up. This isn’t so bad when it stays all in good fun. However, this intense focus on one vector can create real-world polarization. It can devolve into micro-culture wars around what (or who) counts as “in” or “out.” At its extremes, it can lead us to dehumanize those who do not fit our niche definition of “right.”
When this happens with metrics, it is Goodhart’s Law: “When a measure becomes a target, it ceases to be a good measure.” In fiction, it’s Flanderization, the tendency of a few particular traits of a character to become their entire personality over time. When we maximize for one dimension, we tend to end up with totalizing ideologies.
These domain-specific laws are an instance of something deeper: blinders and single goals never work. It may be that an essential element of human consciousness is the ability to balance competing claims on how to decide what is good. When there is no singular right move, we need to apply human judgment. This need for balance holds for everyday concerns such as designing systems: how do we design for engagement and quality? It also holds for the deep values that impact us all: how do we realize justice and mercy?
Why is this need for balance so fundamental? If we can solve a problem with a simple rule — by simply maximizing a single metric — then evolution, whether biological, social, or technological, will eventually generate a solution. Design and ethics both arise from the leftovers, the problems that are unresolvable due to ambiguity, complexity, dilemmas, and the incommensurability of chosen values. These problems of balance are “human-complete” problems.
When it comes to problems of balance, every solution is wrong when taken in isolation, and many (although far from all!) solutions are satisfactory when applied with contextual nuance. The challenge — and the potential — comes from using meta-rational tools to understand when to keep pushing in the same direction and when to start heading another direction before we hit a totalizing extreme.
🛣️🚩 Signposts
Clues that point to where our changing world might lead us.
🚏👩🎤 Grimes says she’ll split royalties with people who use her voice to make AI-generated music
After an AI-generated song that mashed up the styles of Drake and The Weeknd went viral, the music star Grimes announced on Twitter:
I’ll split 50% royalties on any successful AI generated song that uses my voice. Same deal as I would with any artist i collab with. Feel free to use my voice without penalty. I have no label and no legal bindings… I think it’s cool to be fused w a machine and I like the idea of open sourcing all art and killing copyright.
🚏⛲️ A dried-up California lake is reappearing thanks to record flooding
Tulare Lake in central California was the largest freshwater lake west of the Mississippi until the mid-1800s, when farmers started diverting the rivers that fed it; the lake completely dried up by 1900. But California has been pummeled by heavy rain and snow this spring, and the lake has reappeared with a vengeance, swallowing up fields and low-lying towns — it’s said that you can identify drowned roads by the telephone poles that still stick up out of the water.
🚏👻 Snapchat is getting swamped with 1-star reviews as people blast its “My AI” chatbot
In early 2023, Snapchat’s average rating on the iOS App Store sat at 3.05 out of 5 stars; over the last week, its average rating has been just 1.67. The biggest complaint has been with Snapchat’s “My AI” chatbot, which is permanently pinned to the top of a user’s home screen and can only be removed if you pay for a premium Snapchat subscription. Users have criticized the bot as unnecessary, intrusive, and even “creepy”: the bot tracks your location and sometimes comments on it when you chat with it.
🚏🏖️ More than half of all Americans don’t use all their vacation days
According to a new survey, just 48% of US workers use all of their provided vacation days. Part of the reason is the rise of remote work (you can now run errands in the middle of the day without having to take a day off), but another reason is the simple fear that taking vacations might harm one’s career. Many respondents said they “worry they might fall behind” or that their “boss discourages time off.”
🚏🦞 The Pacific garbage patch hosts an ecosystem of coastal species in the middle of the ocean
Marine biologists have been taking a close look at the North Pacific Subtropical Gyre, an infamous 7-million-square-mile patch of plastic trash in the middle of the Pacific Ocean. They’ve found a surprising number of coastal species like mussels, barnacles, and anemones living on the plastic rafts. These creatures need something solid to latch on to, so the open ocean isn’t normally hospitable to them; they’ve been known to raft out to sea on natural objects like tree trunks or seaweed, but those decay or sink too quickly to be a long-term home.
📖⏳ Worth your time
Some especially insightful pieces we’ve read, watched, and listened to recently.
Social Media is Doomed to Die (The Verge) — Observes a consistent pattern that ruins social media products: they realize that “short-form videos from strangers are more compelling” than content from your friends and family, so they morph from social networks into mindless entertainment. “Call it the carcinization of social media, an inevitable outcome for feeds built only around engagement and popularity.”
Whatever Happened to the Starter Home? (New York Times Upshot) — Argues that, as land values in the US have skyrocketed and misguided (though often well-meaning) regulations have increased the costs of construction, it’s no longer economical for many home builders to make cheap “starter homes” for young families. The good news is that many cities are reworking their zoning laws to encourage multi-family housing on former single-family lots, and builders are taking note.
On Extending Human Understanding of Animal Sensory Worlds Through AI (Mystical Silicon) — Inspired by Ed Yong’s latest book, An Immense World, David Gasca observes that, just like animals experience the world in a way that we humans cannot, AIs might do the same. The essay is an engaging summary of Yong’s book, which seems like a worthy read.
Twitter’s New Verification System Is a Complete Nightmare (Mecha-Jesus on Reddit) — A pithy explanation of how Twitter’s new blue-check policy causes adverse selection: the people most likely to buy blue checks are those who crave attention but can’t normally make good enough posts to earn it. And because their comments are pushed to the top of every post, Twitter users can’t ignore them.
The Secret History of the Supernova at the Bottom of the Sea (Nautilus) — By studying ferromanganese crusts growing on the bare bedrock of underwater mountains about 7 million times slower than human fingernails, one physicist is trying to trace supernova explosions through geologic time. The goal is to explore the effects of nearby supernovae on our planet’s climate and see how they indirectly shaped the trajectory of life on Earth.
🔍🌼 Lens of the week
Introducing new ways to see the world and new tools to add to your mental arsenal.
This week’s lens: Intentional farming.
Dandelions are commonly considered weeds because of their ability to grow in a wide variety of environments… including ones where we don’t want them to grow. This is why we use them to illustrate the evolutionary strategy of rapid reproduction with a low survival rate.
But there are two complications to this story of dandelions as unstoppable weeds. For one, even hardy plants can’t grow absolutely everywhere. Whatever we’re growing — be it a dandelion, an elephant, or anything in between — its success is based not only on its inherent fertility, but also on its physical environment. A fertile seed on fertile ground will grow. One on barren, rocky ground will not. As in the parable of the sower, whether or not a seed grows depends as much on where it lands as the seed itself.
A second twist is that we only call things “weeds” if we don’t want them. If we’re creating a meadow of wildflowers, we might consider dandelions highly desirable. (Similarly, European settlers brought the dandelion to North America because it was a fast-growing crop!)
In other words, whether we call something a “success” depends on its social environment. A seed can be desired or it can be unwanted. If it’s wanted, its growth is a success. If it’s unwanted, its growth is a problem. And sometimes, as in the parable of the tares, we cannot tell whether or not a particular seed is desirable or not until it’s had a chance to grow and reveal its shape.
This sets up a nice 2x2. If a desired seed is planted in fertile ground, it’s a success. Planting an unwanted seed in infertile ground is also a type of success, although perhaps a long-shot bet where most seeds will be set up for failure. Anything else, be it a wanted seed in infertile ground or an unwanted seed in fertile ground, is a failure (although each type has its own remedy).
Oftentimes, when we look at a successful or failed project, we tend to look for internal factors to explain the outcome. We assume it is the fault of the seed if it didn’t sprout. We decide that undesirable outcomes come from bad ideas. However, what this lens teaches us is that success or failure is more nuanced. An offhand comment from an executive can lead to an unintentional and unwanted project — a weed. A good idea that fails to thrive might have been planted in the barren soil of an org whose incentive structure doesn’t meet the project’s needs.
This lens also helps us to set up our seeds for success. If we want to plant a large number of seeds so we can harvest the success, we need to provide a fertile environment for them to grow in. We also need to let them grow long enough to reveal whether or not they fit in our social context. If we try to judge the seed solely on its own merits, we’re unlikely to harvest success.
© 2023 The FLUX Collective. All rights reserved. Questions? Contact flux-collective@googlegroups.com.