Episode 128 โ January 11th, 2024 โ Available at read.fluxcollective.org/p/128
Contributors to this issue: Erika Rice Scherpelz, Melanie Kahl, Dart Lindsley, Neel Mehta, Boris Smus, MK
Additional insights from: Ade Oshineye, Ben Mathes, Justin Quimby, Dimitri Glazkov, Alex Komoroske, Robinson Eaton, Spencer Pitman, Julka Almquist, Scott Schaffter, Lisie Lillianfeld, Samuel Arbesman, Jon Lebensold
Weโre 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 weโve noticed in recent weeks.
ย โThe trick to having good ideas is not to sit around in glorious isolation and try to think big thoughts. The trick is to get more parts on the table.โ
โ Steven Johnson
๐ฑ๐ณ Seeds of learning
We often think of development work linearly. When planning we might ask: whatโs the risk? What type of a project is this? How much work? Once weโve categorized the parameters, we send these ideas through structured programs: incubate, accelerate, scale. This structured process can be powerful, but it doesnโt work for all ideas. Providing the right conditions to grow, pouring in resources to accelerate growth, and then spreading the results as widely as possible works well for ideas that are well suited to our current technology, culture, and aspirations. These ideas are like the seeds weโve cultivated for human use over the decades and centuries: theyโre a great fit for exactly the sort of environment weโre asking them to thrive in.
However, some ideas fail to thrive under that model, just as some seeds donโt thrive under our standard regimen of cultivation. When we look to nature, we see that the ways that seeds and spores spread are varied โ their growth isnโt linear or exponential, as we often expect of our ideas. What might we learn from nature about how to support ideas?ย
Seeds and spores are dispersed โ and germination activated โ by wind, water, or heat, or on (and through!) animals. Each needs its own conditions to survive and thrive. Those methods of reproduction co-evolve in the environments they are a part of, but even the most effectively evolved dispersal method also involves a good dose of luck.ย
The fluffy seeds of a milkweed or dandelion can float on the wind or get caught on fur. They are so common as to generally be called weeds. They call to mind memes: common and widespread and individually often quite trivial. But if you look at them as a whole, they are influential and beautiful in their own way.ย
The helicopter of a maple must travel what, for a seed, is a long distance to find a place where it can thrive outside of the literal shadow of its parent. This seed has stronger propellersโcrafted to carry it further away from where it began to find a place where it has room to grow. This calls to mind innovation within institutional settings. New ideas can die hidden away in institutional settings due to the pressures of the innovatorโs dilemma. Such ideas may need to helicopter away from their parents to thrive. Many startups originated as helicopter seeds from established companies.ย
Other plants, including fire followers, like the mariposa lily or the serotinous cones of the yellow pine, need fire. The highest quality, richest environments do not help those seeds thrive; they come to life in the extremes that kill others. They have evolved to thrive in heat โ or survive it. Some ideas are similar, like the mRNA vaccines whose history goes back decades before COVID-19 provided the incentive to turn them into a practical life-saving technology (one that continues to find more applications). Some ideas need a sense of urgency to get them over that initial hump from idea to application.
Some species of moss are specialized to very particular environments, such as the droppings of a particular deer. Try to move these spores to another environment, even one that seems similar, and they will fail to take. Some ideas are like this. Within a particular community an idea may spread, but take it out and it sputters out. Specialized communities, be they centered around hobbies, beliefs, or fandoms, often have this nature.ย
Other seeds travel by way of wandering animals. They make themselves tasty enough to eat (endozoochory) or are sticky enough to hold onto feathers or fur (epizoochory). Wrapped in fruit, seeds become attractive and nourishing โ and their readiness for germination is often aided by digestive juices and built-in fertilizer. In forests, especially jungles, this method of dispersal is critical, driving ecological health and co-evolution. Losing these creatures that digest and share nourishing seeds โ or ideas โ may threaten the very health of an ecosystem. (Or the opposite: a long-distance dispersal of a seed hitching a ride in a sheepโs fleece may spread an invasive species.)
Seeds are packets of potential, concentrated energy waiting on the right environment to be released. Just as shapes of ideas are specialized to be able to best take advantage of specific environments, seeds have co-evolved to thrive in different conditions. And while no seed or spore can thrive in every environment, almost every environment is able to be colonized by some organism. So when we see an environment where our brilliant innovations arenโt sticking, instead of asking ourselves, โwhy isnโt this idea thriving here?,โ we can turn the question around and ask, โWhat sort of insights would thrive here?โ Letting the environment guide us may make us feel less brilliant, but it is also likely to help us be more effective.
Want to learn more? Go down the ๐ Wikipedia rabbit hole, ๐ผ peruse this exhibit, or check out ๐ this childrenโs bookโฆ. or ๐พ simply notice on your next walk.
๐ฃ๏ธ๐ฉ Signpostsย
Clues that point to where our changing world might lead us.
๐๐ The SECโs Twitter was hacked to โannounceโ Bitcoin ETFs becoming legal
The Security and Exchange Commissionโs account on Twitter/X was compromised and used to post a message saying that Bitcoin spot ETFs (a long-awaited financial product in the crypto world) had been approved; the SEC swiftly regained control and took down the tweet. Finance writer Matt Levine wrote that this might not have been a true attempt at market manipulation: everyone knew ETF approval was coming that same day, so any news had already been priced in. Rather, it might have just been an attempt to โtrollโ the SEC by making their account announce the approval, retract the announcement, and then have to re-announce it the next day.
๐๐ Hawaii replaced its last coal power plant with a huge battery
Some places keep fossil fuel plants open because, if nothing else, they can always generate power when you need it (whereas renewable energy can be intermittent). But Hawaii recently replaced its last coal power plant with an array of 158 Tesla Megapacks, which can discharge power into the grid within just 250 milliseconds of getting a request (even faster than the old coal plant could!). The batteries will store excess renewable energy produced during the day (preventing green power from getting โcurtailedโ because nobody needs it at that moment) and release it during the high-demand evening hours.
๐๐ค The actorsโ guild signed a deal to allow AI voiceovers in video games
SAG-AFTRA, the American actorsโ union, signed a deal with an AI voiceover company that would allow the use of AI to recreate the voices of real performers, though not to create synthetic voices. Voice actors need to give their consent before their voice is used in this way, and they need to be compensated too. (The unionโs president called the deal โa great example of AI being done right,โ but many voice actors were unhappy, calling the deal a โmistakeโ or โgarbage,โ and asking how it couldโve been approved without being put to a vote.)
๐๐ A new โmouse for your mouthโ lets you type and scroll using your tongue
A โretainer-like trackpadโ unveiled at this yearโs CES lets you use your tongue as a Bluetooth input device for your phone or computer. By moving your tongue around on the disc, which sits on the roof of your mouth, you can scroll and type; you can also press on the left or right side to โclickโ or trigger other hotkeys. The gadgetโs creators say it was designed to help people with disabilities like paralysis, as well as professionals who use their hands a lot, like mechanics and surgeons.
๐โณ Worth your time
Some especially insightful pieces weโve read, watched, and listened to recently.
New Yearโs Message: Moving Fast and Breaking Things Is the Opposite of Tech Optimism (Techdirt) โ Disagrees with some prominent โtechno-optimistsโ who argue that putting safeguards in place around new technologists will hinder progress. The author argues that the opposite is in fact true: โmoving fast and breaking thingsโ causes backlash that ultimately slows down adoption. If you deploy new tech thoughtfully, you can avoid the regulations and public enmity that could otherwise derail it.
Normcore LLM Reads (Vicki Boykis) โ A wide-ranging and โanti-hypeโ list of technical deep-dives into large language models, including the history of NLP and language models, foundational concepts like deep learning and transformers, how LLMs are designed and trained, the mechanics of RLHF and fine-tuning, and the UX of building products around LLMs.
How Terrorism Ends: Understanding the Decline and Demise of Terrorist Campaigns (Middle East Forum) โ A review of a book by Audrey Kurth Cronin that observes that terrorist groups often refuse to negotiate and virtually never attain their given political aims. Their terrorism often undercuts their stated aim, although it does boost internal cohesion. The way these groups typically fail is via backlash against the gory violence itself, something thatโs especially relevant in todayโs Middle East.
Trading Places: The City and the Suburb (Alex Marshall) โ Observes that, because economic activity has largely moved from city centers to suburban office parks, cities and suburbs have swapped their traditional roles in economic life. Suburbs are now โsoullessโ industrial areas of โuncontrolled technological forces,โ while cities represent โstability, community, and the human scale.โ Many downtowns have become tourist-friendly โantiquesโ rather than wealth producing engines โย echoing McLuhan, who wrote that โthe sloughed-off environment becomes a work of art in the new invisible environment.โ
๐๐ Lens of the week
Introducing new ways to see the world and new tools to add to your mental arsenal.
This weekโs lens: Newtonian wrongness.
Traveling often comes with moments of disorientation, where you think you are in one place but youโre actually not. Sometimes you find that you're completely off base: you thought you were going west but you're actually heading south. Other times, youโre off but directionally correct: you thought you had gone four blocks, but it turns out that youโd only gone three.
The first case is being wrong in the traditional sense: youโre wrong in a way that is definitely not right. The second case might be called Newtonian wrongness. Classical Newtonian mechanics are, strictly speaking, wrong. They break down at extreme scales, such as moving near the speed of light. However, at everyday scales, the errors of applying the โwrongโ models of classical mechanics don't make any practical difference. Similarly, if youโre going in the right direction, it might not matter if you donโt know exactly where you are as long as you know the big pieces like where to turn.
Newtonian wrongness can help us decide how to react when we find ourselves disagreeing with others. Is the disagreement fundamental? For example, will the result of the discussion end up taking a product in what we see as a problematic direction? Then itโs probably worth voicing our concerns and aiming for alignment. But maybe the disagreement is just one of Newtonian wrongness. For example, the detailed definition of metrics matters, but, whether we measure daily active users (DAU) day-by-day or as a rolling weekly average, directionally both are similar.ย
When applying the lens of Newtonian wrongness, itโs important to be aware of what scale youโre working at. That same โtrivialโ difference between different ways of measuring DAU do matter to the person who is implementing them or interpreting them in a graph. The takeaway isnโt โsome differences donโt matterโ but a more relativistic question: does this difference matter for the larger question at hand?ย
ยฉ 2024 The FLUX Collective. All rights reserved. Questions? Contact flux-collective@googlegroups.com.