🌀🗞 The FLUX Review, Ep. 116
September 7th, 2023
Episode 116 — September 7th, 2023 — Available at read.fluxcollective.org/p/116
Contributors to this issue: Boris Smus, Neel Mehta, Dart Lindsley, Erika Rice Scherpelz, Dimitri Glazkov, MK
Additional insights from: Ade Oshineye, Gordon Brander, Stefano Mazzocchi, Ben Mathes, Justin Quimby, Alex Komoroske, Robinson Eaton, Spencer Pitman, Julka Almquist, Scott Schaffter, Lisie Lillianfeld, Samuel Arbesman, 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.
“If we want things to stay as they are, things will have to change.”
— Giuseppe Tomasi di Lampedusa
💃📸 Dancing with art
What makes a painting good? What makes a photograph good? What makes an image created with generative AI good? Part of the answer is these questions don't have a singular answer. As our medium of expression changes, so does our sense of what makes something meaningful, beautiful, or moving.
Photography reduced the value of hyperrealistic paintings and opened the door to more abstract forms of expression. This isn't because people no longer value realistic expression. Rather, that need shifted to a medium better suited for it, allowing the previous medium to explore new paths.
This difference in how we evaluate a work relates to our perception of its creation and its curation.
When something is hard to create, the bottleneck is, “How do we get more?” In such an environment, the capacity to create is valuable in itself. Before photography, a mediocre portrait painter could thrive in any town—there were always unmet needs for portraits. However, over time, the cost of creation tends to decrease. Scribes aren't necessary in a society where writing is ubiquitous. Portrait painters don't need to capture memories when a camera is always in your pocket.
Does that mean an object easily created loses all value? The long path from the invention of photography to its recognition as an art form shows that skill is key in how we evaluate value. That same path also shows that, while the demonstration of skill is constant, the nature of that skill changes.
This is where the idea of curation comes in. It turns out that there is nothing inherent in one brushstroke which makes it more valuable than another brushstroke… or than a pixel. Expression is meaningful within a feedback loop of experimentation and evaluation. Whatever the medium, the history of art is the expression of new ideas and reinterpretation of old ones. Many experiments fail. Even those that succeed become commonplace as others learn to copy them. In this copying, we may find a new adjacent possible, leading to new spaces to experiment.
The introduction of photography, its evolution into the ubiquitous cameras on our mobile devices, and now generative AI have turned what was once a bottleneck into a flood. Even before Stable Diffusion, Midjourney, and other generative AI tools, humanity was creating more imagery in a moment than in years of the past. Both the ease of creation and the number of creators have accelerated rapidly.
We now live in a post-scarcity era for content. We have shifted from a creation problem to a selection problem. If the question for a creation problem is “How do we get more?,” then the question for a selection problem is “How can we tell what is good?” We can try to hold onto the past, trying and failing to apply the same criteria at an incomparable scale. Alternatively, we can accept that what “good” means is an ever-evolving dance. It is our participation in that dance, both as creators and consumers, that creates meaning.
🛣️🚩 Signposts
Clues that point to where our changing world might lead us.
🚏🧑🎨 An “AI” startup for generating 3D models used outsourced human labor instead
One buzzy AI startup drew attention for its machine learning technology that it claimed could automatically convert 2D illustrations into 3D models. But a recent exposé revealed that the company’s AI produced poor quality outputs, and thus the startup relied heavily on human artists to build the models; the artists sometimes even created the 3D models from scratch, with no AI assistance. The artists were hired to “generate low quality 3D assets quickly” in as little as 15 minutes, and they were paid $1 to $4 per model they churned out.
🚏⛽️ China is set to hit peak gas this year as EV adoption soars
Sinopec, China's largest fuel distributor, recently announced that it expects gasoline demand in China to peak this year, two years ahead of previous projections, largely due to the rapid adoption of electric vehicles. China has been a significant driver for global petroleum demand, but with EVs accounting for 38% of new passenger-vehicle sales and ride-hailing fleets increasingly going electric, the landscape appears to be shifting.
🚏🗞️ Gizmodo fired its Spanish staff and replaced them with an AI auto-translator
G/O Media, the parent company of the technology publication Gizmodo, abruptly fired the staff of Gizmodo’s Spanish-language site and began replacing their work with AI auto-translations of Gizmodo’s English articles. The Spanish staffers, in addition to hand-translating English pieces, also wrote original content catering to Spanish-speaking readers; that original content will seemingly be discontinued as a result of the shift. Some observers have noticed glitches with Gizmodo’s auto-translation system, like when it once switched from Spanish to English in the middle of an article.
🚏🇦🇮 Anguilla could earn $30 million this year from its .ai domain names
Anguilla, a tiny British island territory in the Caribbean, controls the “.ai” top-level domain (TLD), just as, for instance, “.in” was assigned to India and “.fr” to France. As the AI space has exploded, so too has interest in “.ai” domains. Anguilla is projected to earn up to $30 million this year on domain registration fees; that alone would account for 10% of the island’s GDP. It’s an unexpected boon for the territory, which has seen its tourism-based economy suffer ever since COVID hit.
📖⏳ Worth your time
Some especially insightful pieces we’ve read, watched, and listened to recently.
Enfeebling Supports (Atoms vs. Bits) — Introduces the concept of “enfeebling supports,” which are crutches that provide immediate comfort or assistance but ultimately weaken your abilities in the long run. “Strengthening supports,” in contrast, not only help in the short term but also improve your ability to perform a task unaided. Examples of enfeebling supports include padded shoes and overprotective parenting, while strengthening supports include pull-up assist bands and weightlifting shoes.
Startups: The Door Is Closing (European Straits) — Echoes geopolitical strategist Peter Zeihan’s view that young talent and abundant capital are necessary for innovation; as both of those resources become scarcer across the world, large organizations will gain an advantage over startups. The coming decade of business may thus come to resemble the 1970s: execution will be favored over innovation (as can be seen with the rise of B2B SaaS startups), consolidation will rule the day, and VCs may need to cede ground to the more traditional business playbooks of private equity.
The Eros Monster (Harper’s Magazine) — A well-written chronicle on hopeless, unrequited love, with Agnes Callard’s philosophical musings and open questions sprinkled throughout, such as why people “lock themselves into dyads of exploitative misery.”
The Mountains at the Bottom of the World, Part 2 (Everything is Amazing) — Takes us on a tour of the undersea Mid-Atlantic Range (the planet’s longest mountain chain!) and a field of hydrothermal vents fueled by a unique chemical mechanism. The fact that life can thrive near these vents, ultimately fueled by chemicals rather than the Sun, indicates that life might just be possible in far-flung spots like the icy oceans of Jupiter’s moon Europa.
🌀🖋 More from FLUXers
Highlighting independent publications from FLUX contributors.
FLUX’s own Dimitri Glazkov recently published We Are Entering a Maker Renaissance in the popular newsletter Every.to. In his essay, Dimitri examines the conditions that typically give rise to inventors and entrepreneurs, and he concludes that generative AI is fueling a new “makers’ era.” Along the way, he shares some great insights about technological innovation and competition, including:
“A good marker of a wide-open space is that a typical market-sizing exercise keeps collapsing into a fractal mess, creating more questions than answers. It’s not just one thing that suddenly becomes possible, but a whole bunch of things—and there’s this feeling that we have only scratched the surface.”
🔍📆 Lens of the week
Introducing new ways to see the world and new tools to add to your mental arsenal.
This week’s lens: Peanut buttering.
It’s time for quarterly planning! And, as usual, our list of things we want to do exceeds our ability to do all of them. So, what do we do?
A common temptation is to engage in peanut buttering, where we take our available resources and spread them out over a large surface area. In project planning, this might mean spreading people over many small projects. In urban planning, it could manifest as tiny pedestrian safety improvements for all intersections.
As commonly used, peanut buttering suggests wasted opportunities. Instead of making meaningful progress on a headline issue, we make incremental progress on a much broader set of less important concerns. In situations where there is a clear prioritization, it’s generally better to invest the maximal useful amount of resources into the top priority. If we still have the capacity to do more, go to the next most important thing on the list and repeat.
But is peanut buttering always bad? Taken to an extreme, where we invest in everything and progress on nothing, yes. However, there is a middle ground which is often described by another food metaphor: not putting all of our eggs in one basket. In a situation where priorities are not clear, it can be useful to allocate our resources to create a balanced portfolio of risks and rewards.
We don’t want to spread our resources so thin that progress becomes impossible. Nor should we concentrate them to the point where a loss is devastating. By looking at the underlying risk/reward profile of the adventure ahead, we can avoid either extreme. We can give more resources to that which is more important while not completely neglecting that which might help us in a pinch.
© 2023 The FLUX Collective. All rights reserved. Questions? Contact flux-collective@googlegroups.com.