🌀🗞 The FLUX Review, Ep. 158

August 15th, 2024

Aug 16, 2024

Episode 158 — August 15th, 2024 — Available at read.fluxcollective.org/p/158

Contributors to this issue: Ade Oshineye, Ben Mathes, Erika Rice Scherpelz, Neel Mehta, Boris Smus, Dimitri Glazkov, MK

Additional insights from: Justin Quimby, Alex Komoroske, Robinson Eaton, Spencer Pitman, Julka Almquist, Scott Schaffter, Lisie Lillianfeld, Samuel Arbesman, Dart Lindsley, Jon Lebensold, Melanie Kahl, Kamran Hakiman, Chris Butler

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 best and noblest gifts of humanity cannot be the monopoly of a particular race or country; its scope may not be limited nor may it be regarded as the miser’s hoard buried underground.”
― Jawaharlal Nehru

🎱🕺 Logic and gut

We learn early in life that logic is the preferred way of sensemaking. We’re taught that the rigorous application of mental energy to arrive at a conclusion, inference by inference, is the superior way to organize our minds. In this way, logical reasoning is the foundational technology, on top of which all other scientific and technological progress resides.

It’s also a relatively recent discovery. To the best of our knowledge, we humans have spent many thousands of years relying on our gut instincts. Gut instinct still has a place in our lives. This older way of sensemaking can be robust and reliable amid disaster when we’re forced to act first and think later. It can alert us to situations where explicit reasoning is missing something. When misused, it can lead to cognitive biases, where we utilize shortcuts in situations where they do not apply.

The border between the gut and logic is porous and fuzzy. It can be challenging to tell if we’re firmly on one side of it or have crossed to the other. We can explore this challenge with our favorite ‘lensical’ device: a 2x2.

The horizontal axis delineates the border between logic and gut as it occurs in a given situation: what we actually do. The vertical axis is our perception, on which we rely. The quadrants formed by the two axes reveal the thinking we’re engaging in vs. what we believe we’re doing. The top-left and bottom-right quadrants are relatively boring: our introspective selves are in alignment. We are either acting on a gut feeling or being entirely cerebral, knowing precisely what’s happening.

In the top-right quadrant, we believe we’re using a gut instinct but are actually driven by logic. Although this may seem unintuitive, it’s a fairly common spot for many “type A” folks: we’re asked to act spontaneously, or be “cool”, and so we try — all the while performing intense computations in our heads on what exactly that should look like.

The bottom-left quadrant is the opposite: we believe we’re thinking logically but are driven by our gut. The classic “What the heck was I thinking?!” is the typical reveal of this dichotomy, and it usually happens after the fact.

In environments where logical reasoning is encouraged, be wary of the bottom-left space. Although gut and logic seem easy to distinguish, it can be difficult to tell what’s happening inside our minds. When we’re triggered, exhausted, or stressed, we can sound perfectly reasonable to ourselves while making giant leaps in our assumptions and surfing the high emotional seas.

There’s no easy way to detect this state, but a great trick is to slow down and sit with the problem. Our gut thinking’s temporal nature reveals itself more clearly with time. This lets us find our way back to the land of reason and use those signals from the gut as an input, not the driver.

🛣️🚩 Signposts

Clues that point to where our changing world might lead us.

🚏🔌 More than half of new cars sold in China were electric or hybrid

This July, 51% of all new cars sold in China were fully electric or hybrid, up from just 36% a year ago. It appears that battery-only cars edged out hybrids, with battery cars grabbing 28% of the overall market share (compared to, presumably, 23% for hybrids). Meanwhile, the combined market penetration of EVs and hybrids has been hovering at around 18% in the US, although the sector has enjoyed impressive growth over the last five years.

🚏🐍 An AI research system modified its own code and tried to run itself

One AI research firm has developed an AI product that claims to be able to use LLMs to produce novel machine learning research papers, from generating ideas to writing code to actually writing the paper. The company was surprised to find that its AI started modifying its own code to help it solve problems. In one case, the AI’s experiment took too long to complete, so it edited its own Python code to give itself a longer timeout window. Elsewhere, it (for some reason) edited a Python script to add a system call that’d rerun the very same script (which, predictably, caused an infinite loop).

🚏🔐 NIST has released three post-quantum encryption algorithms

It’s well-known that once sufficiently scaled up, quantum computers can break most modern encryption algorithms. So, the US’s National Institute of Standards and Technology (NIST) has been keenly working on developing quantum-resistant (or ‘post-quantum’) encryption methods. This week, NIST released three much-awaited post-quantum cryptographic algorithms, with a fourth coming later this year. NIST says developers can start using these new algos straight away.

📖⏳ Worth your time

Some especially insightful pieces we’ve read, watched, and listened to recently.

Cannibal AIs Could Risk Digital “Mad Cow Disease” Without Fresh Data (Science Alert) — Observes that AIs trained on synthetic data become corrupted and start generating progressively stranger and more homogenous outputs, drawing an analogy to the disease that afflicts cows who were fed the infected remains of other cows. (It’s similar to the concept of ‘Habsburg AI,’ which also warns of the risks of model “inbreeding.”) Closed feedback loops hurt AI, and the only cure is to inject “fresh” new training data.
Why Western Designs Fail in Developing Countries (Design Theory) — Tells the story of the ill-fated One Laptop Per Child project: its laptop looked like a good idea to donors, but in practice, students had no context about what a laptop could do and no infrastructure to actually use it. Lower-tech solutions that solve a more fundamental need tend to do better: consider a skylight made of a water-filled plastic bottle, which brings more light into the house and gives students more time to study.
Ketman: Living in Disguise to Gain Acceptance (Ezega) — Deriving from the Arabic word for discretion and concealment, the habit of Ketman encourages one to suppress their deepest beliefs in their professional or creative pursuits so that the individual can be accepted and ultimately thrive. Ketman can also be a source of pride, where a “believer raises himself to a permanent state of superiority over the man he deceives.”
5G Was an Over-Hyped Dud. Prepare for Nobody to Learn Absolutely Anything From the Experience (Techdirt) — Argues that “5G was an incremental technical improvement that companies tried to tell us was a revolutionary leap,” and in this way it’s a classic example of the tech hype cycle: many new technologies do bring modest productivity improvements, but that’s not sexy, so people trying to get attention need to advertise potential use cases that may not actually materialize.

💊🍀 A dose of hopepunk

An optimistic sign that our world’s systems are changing for the better.

Baby beavers (“kits”) have been born in urban London for the first time in 400 years thanks to the Ealing beaver project, which introduced a family of the critters to a nature reserve in west London. The project aims to educate the public about the ecosystem benefits that beavers and their dams bring: “improved water quality, reduced flooding, more insects and other wildlife on site, habitat improvements.” And, of course, the staff were delighted to see two baby beavers as a bonus.

🔍🪞 Lens of the week

Introducing new ways to see the world and new tools to add to your mental arsenal.

This week’s lens: Conway’s mirror.

Imagine an organization that produces many proposals that never get implemented, never get approved, and perhaps never even get read. It’s not that the organization is doing nothing. They’re just not doing what their proposals say.

Conway’s law tells us that an organization’s structure impacts the way it operates and the products it produces. Conway’s mirror builds on this idea: the gap between lofty ideas and realized artifacts of an organization can be used to understand when the aspirational — and sometimes even official — organizational structure is not aligned with the effective structure and shadow processes that tend to be followed in practice.

An organization's idea artifacts—the proposals, presentations, and quarterly plans—tend to reflect its aspirational stance. The actions, on the other hand, reflect the actual values, beliefs, and priorities of the people who created them.

How artifacts and actions relate helps us understand organizational (mis)alignment. When we see a gap, we know that something is off. Like a mirror, the outcomes of our actions could reflect that our artifacts are too flat or slow-moving for reality. They could tell us that we are enhancing our ability to have an impact by adapting to changing conditions. Or, perhaps, they could signal a lack of acceptance of a change. Two teams that resent being reorged together may reflect the new, unified, homogenous org structure in their plans, but in practice, they bicker, fight, and never get anything done. At the extreme, Conway’s mirror can warp into Conway’s funhouse mirror, a form of kayfabe where the distortion between the artifacts and the action is intentional.

Conway's mirror can help us see the outcomes of our organizations in a new light. By examining the artifacts that we produce and holding them next to the mirror of the outcomes, we can gain insights into our own values, beliefs, and priorities — and whether or not they are helping us go in the desired direction or preventing us from getting there. At the same time, Conway’s mirror can help us see the present reality more clearly. It acts as a true mirror for our organizations, showing us what we look like to others, not just what we are used to seeing.

🌀🗞 The FLUX Review