My Post-AI Writing

March 8, 2024 By

I was asked in a DM conversation whether I use AI for writing, and I said no, it would be like going for a walk in my car. The only people who seem to directly use AI for writing are people who don’t write for pleasure, but have to write a lot of functional, business-like things that are too varied for boilerplate templates. The sort of writing that might require intelligence or expertise, but which doesn’t offer much pleasure or creative challenge to the writer. It’s an instrumental sort of writing.

Which is not to say AIs can’t do creative writing. I’m only saying that people who enjoy creative writing purely or mostly for the process itself have no real use for AI assistance. To a lesser extent this seems true of coding as well. A lot of the use I’ve seen is for the tedious bits. And as with writing, that’s not to say AIs can’t do creative coding requiring insight. But people who code for fun probably have less use for it, though I suspect “copiloting” is likely more fun with code than writing.

The more the ends matter more than the means, the more AI is helpful. If you like flying a plane, a copilot just cuts in on your time at the controls. If you’re just trying to fly somewhere, you’re happy to let the copilot fly. For me the “ends” of writing barely matter at all. It’s all about the means.

Can AIs enhance creative satisfaction of exercising the means? I imagine so. I can imagine an assistant that I explain my goals to, and it acts like an improv partner, perhaps writing every other sentence. I suspect ChatGPT can already do this well. But that’s a different creative process I’d have to learn to derive satisfaction from, like learning to go rowing as a substitute for going on walks.

But AI has had an indirect positive and enjoyable effect on my writing: It has made me lower my craftsmanship standards, which were never very high to begin with. This is one reason I’m writing a lot more this year. The causal chain from AI is subtle, and AI is not the whole explanation for changes in my writing, so let me try to unpack the part that is.

[Read more…]

Matter and Life

March 7, 2024 By

Two articles about matter and life have been on my mind for a while. The first is Life Helps Make Almost Half of All Minerals on Earth, in Quanta magazine. The second is this article in Nature, Global human-made mass exceeds all living biomass.

We rarely think about matter as a first-class subject of philosophical inquiry, the way we do space and time. Or about the divide between non-living and living matter except through the lens of the presumptive specialness of living matter. Physics, in a way, has been a centuries-long exercise in trying to subordinate matter to time and space (and more recently, waves, fields, and information which are all spatio-temporal flavored abstractions). Non-living matter, of course, likely comprises the vast majority of all matter, even if the universe is teeming with life. But life likely accounts for the vast majority of the variety in matter. Stars make all the heavier atoms when they go supernova (a fact that some people make way too much of). But then when you get to molecules, I suspect most molecules are either themselves organic, or have their macro-structures shaped by life. The Quanta article is about minerals and rocks, which we don’t think of as organic matter — and they’re not in compositional terms, since they’re not complex carbon compounds — but in a procedural sense, which seems more fundamental, they might be. And we’re not just talking simpler examples like sand being made of sea-shell powder or oil from ancient marine organisms. It looks like lots of rocks are shaped by life. Basically half of the material variety you see around you is connected to life and its cycles.

The Nature article takes it one level of abstraction higher. The built environment is also shaped by life. We normally focus on the spatio-temporal; aspect, the geometries and the lifecycles, but consider the material aspect of the majority of the built-environment mass:

  1. Concrete
  2. Reinforced concrete
  3. Asphalt
  4. Glass
  5. Metals and alloys

None of these materials are “non-living” per se. Some, like metals, wouldn’t exist in pure-ish form in an oxygen atmosphere without life. We find pure-ish metals in meteorites of course, but on a planet, except for a few precious metals like gold and silver, pure metals don’t exist. And even the precious ones don’t occur as bulk masses. Arguably reinforced concrete is an “organic” material — a composite that wouldn’t arise in matter without life.

Speaking of oxygen, that too of course, only exists in free form because of life. Several natural cycles of simpler atoms — oxygen, nitrogen, carbon, phosphorus, calcium — are pumped by life. Without life, the cycles of these atoms would be pretty dead (heh!)

Of course, pure non-living processes help. Life as we know it can’t form without tectonic activity starting the churn.

The point of this discussion is that where life exists (based on our n=1 case) living matter is a significant phase in the cycles of all matter, perhaps even the dominant phase. We often talk of life as though it’s a fragile bit of material poetry that is alienated from, and in thrall to, the far vaster non-living processes of matter, but it isn’t. We talk as though the processes of life are less powerful than those of non-living matter, such as wind, rain, or earthquakes. They’re not. Life punches in the same weight class as non-life where the two touch. We should think of “life” as a material-energetic-spatial-temporal phenomenon with the same sort of raw, irresistible power as wind or waves. We say earth to earth, ashes to ashes, dust to dust, but we could say, with equal justice, life to life as the karmic destiny of non-living matter. Under the right conditions, non-living matter can no more resist the irresistible tug of life than living matter can resist death. We lament that there have been half-a-dozen mass extinctions on the planet. We might equally say, there have been half-a-dozen mass vivifications of non-living matter.

Yes, “life” is a lower entropy (but higher-information) more complex emergent state, so non-living matter is in some broad sense “lower,” but the Earth, bathing in the rays of the sun, is not a closed thermodynamic system. Entropy-increase is not the defining feature of life and non-life on Earth.

This macro-scale balance of power between the forces of life and non-life seems to me at least as important as the specific details that define the boundary, like the structure of viruses (which are between living and non-living). I suspect we need a science of macro-scale life comparable to the science of weather. It will be to virology what weather is to basic statistical mechanics.

There are two practical reasons to be interested in the philosophy of matter and life today. The first is climate change. The second is computing.

We talk of the Anthropocene as though life reshaping non-life is “new” but it is as old as geology; as old as sand and free oxygen. We have to get our terms of reference right to think straight about climate action, or as I prefer to think of it, terraforming 101. What’s new is life consciously embracing its material nature and its entanglement with the non-living beyond the mere cosmetic layers of “built” environments. Our material natures run deeper than our inclination to build things.

The materiality of computing is perhaps even more consequential. Computers are, as the line goes, “rocks that we zapped with lightning and tricked into thinking.” I find the fact that you can skip life altogether and make “thinking” a property of all matter — living or non-living — to be more interesting than the fact that the “thinking” can be “intelligent.” It seems to take life to get matter to “thinking” states but this is just a limited perspective. This is more like the fact that it takes life to produce free oxygen than it is a fact about technology and invention. Non-living processes first produced life that produced thought, but that thinking in turn gave rise to new kinds of rocks that can think.

But more than these practical concerns, I think there is a deeper point here about the materiality of life and the vivifiability of matter. Our estrangement from non-living matter, with which we are so obviously entangled at all scales by powerful processes, is the perhaps the last bit of our anthropocentrism (or biocentrism rather) that remains to be broken down before we can truly feel at home in the universe. We are not in the universe, we are the universe.

We’ve already embraced our spatiotemporal nature. We aren’t in space, we are space. Our bodies have volumes; they have extent. Crushed too close together, bodies flow like fluids. That’s what a “stampede” is. We don’t exist in time, we are time. Memories are life and those are just another face of time.

Curiously, it’s been harder for us to achieve this sort of explicit, conscious awareness of our material natures. Many science fiction writers handle space and time well. Few handle matter well (Iain M. Banks and J. G. Ballard come to mind). Many artists do amazing things with space and time through audiovisual media. Materiality has been a harder dimension of life to explore. What’s the Mona Lisa of density? The Beethoven’s Fifth of porosity? (I find most modern explorations of materiality in the fine arts rather underwhelming compared to explorations of space and time). Much of our sense of materiality has remained at the level of craft (weaving, pottery, metalwork) and unexamined felt experiences (have you ever wondered much about “solidity” the way you’ve likely wondered about “time”?). Matter is hard to intellectualize, perhaps because variety is its very essence, especially when life gets involved. Platonic geometry is a very successful philosophical abstraction of space. Music is a very successful philosophical abstraction of time. But earth, fire, water, wind, ether kinda misses the point of matter. The point of matter is variety. Which is, in some ways, the antithesis of abstraction. To try identify 5 “basic” substances is to miss the point.

And life is perhaps just matter’s way of creating more varieties of matter, which makes it even harder to think about, because that just seems to make the variety explode even faster.

Civilizational Functionalism

March 5, 2024 By

For much of history, the grand narrative of civilization, such as it is, has been viewed in terms of capital-P Purpose. Life on earth supposedly has a Purpose. Until modernity, that Purpose was generally understood in religious terms, which meant it was understood in incomplete-by-design ways, with the most important bits of the Grand Design driving the universe to strive towards that Purpose being hidden from view beyond the veil of death. We could only guess at it through mythologies of afterlives, karmic cycling, Judgment Days, and so on. Your part of the Divine Design was to serve the Purpose by living out a small-p purpose driven life, with the two being connected by a shared positive-valence quality called Good/good. You served the Greater Good of the Grand Design by striving towards your small-p purposes doing small-g good, which fed the larger Purpose. Doing so generated Meaning and meaning.

This is all conveniently unfalsifiable, self-serving, and self-soothing. Which is one reason that when humanity discovered Science! it took care to retain that unfalsifiability in new Purpose narratives that rejoiced under a new label: Progress.

I find this whole scheme rather an insult to the intelligence. Is there a better way to understand whatever larger-scale coherence civilization has, and how it aligns with any small-scale coherence we may experience in our lives and actions? I think there is, and the key to it is the word function.

[Read more…]

Poison-Depilling Problems

March 2, 2024 By

Occasionally, we all find ourselves attracted to what we might think of as poison-pilled terms (by analogy to poison-pill clauses in contracts or pieces of legislation). Terms that point to interesting and useful ideas but fatally compromised by a) political baggage, b) unsound analytical provenance, or c) plain distastefulness of associations.

Subjectively, a poison-pilled term feels like the right term for a thing, coined in the wrong place, by the wrong person, for the wrong reasons. Where wrongness is of course relative to a presumption of our own rightness.

Assuming we are not sufficiently persuaded by the power of the term to change sides on the underlying issue, or operating definitions of rightness/wrongness (that would be a powerful term indeed), we are then faced with a poison-depilling problem.

[Read more…]

Economics Memes

February 23, 2024 By

I can’t pretend to understand economics the way economists pretend to since I didn’t go to acting or theology school, but I try to form mental models about economics starting with sociological and technological first principles, which I consider the twin phenomenological grounds economics rests on, and the twin sources of the regulatory epistemic pressures that constrains the more cancerous tendencies of economics as a discipline. History used to be a third source but it’s become too captured by economics now to be useful. A good example of sociology-first economics is Donald Mackenzie’s “engine, not a camera” mental model. A good example of technology-first economics is Moore’s Law.

In the world-view of economists, starting with sociology is exactly the wrong thing to do, since in economics dogma, sociology is just a lot of narrative superstition. This is one of the best reasons to start with sociology. Unsurprisingly, the parts of sociology that have suffered the most from the replication crisis are the parts that look the most like economics, based on shaky statistical reasoning on questions that should probably have been only explored with more conservative ethnographic methods to begin with. And when economists do incorporate sociological thinking at all, it is in the form of, well, mathematized narrative superstition a la behavioral economics, which carries over a lot of the irreproducibility sociology and social psychology have become associated with. In the sociology-economics relationship, economics makes a good servant but a poor master. When economics bosses sociology around, both get worse.

Economists tend to be less willing to challenge technology narrative forcing functions. I’ll leave that relationship for another day. Let’s talk about how sociology can remain the master of economics: Economics memes.

[Read more…]

Four Modes of BDFxing

February 21, 2024 By
This entry is part 2 of 2 in the series Lunchtime Leadership

Made up a 2×2 after a long time. And am kicking off a new Lunchtime Leadership blogchain. Name inspired by Hitchhiker’s Guide joke about most work being done by random people who wander into offices while leaders are out to lunch, see something worth doing, and do it. BDFxing, the idea I introduced in the first part, is of course a core idea. It stands for Benevolent Dictator for x, where x is a time period between an hour (a meeting) to about a year. Happily it could also stand for eXecution, since usually it is execution needs that create leadership needs. The central dogma of Lunchtime Leadership is that most things don’t need leadership most of the time. BDFxing is one obvious implication of the dogma — that most leadership should be part time. Being a leader is fun. Just not 24 hours a day, 7 days a week, 365 days a year. Most of the time, for most things, there should be no leaders, and when there is, it should be a focused leader suited to the situation, who should only lead while the need exists. With a norm of people just walking away if they think somebody is trying to lead unnecessarily, which makes BDFxing a self-limiting pattern that resists the power-hungry. So the follower discipline corresponding to BDFxing is opt-in followership. If you don’t think something needs leadership, don’t follow. And everybody should be capable of stepping up to lead some of the time, in some way. A lot of my thinking here is shaped by the Yak Collective, which runs entirely on BDFxing and opt-in followership.

The “size” of a leadership episode could be measured as a product of stakes and energy. Ie, the value of what’s at stake times the energy output rate (in the sense of Andy Grove’s idea of “high-output management”) that needs to be put in for the duration, to make something happen. Here’s the 2×2 with the resulting for BDFxing archetypes (which are transient roles, not personalities).

Obviously, high-stakes, high-energy BDFxing makes for the shortest bursts of leadership, and the typical situation is when something that is the output of a longer period of leaderless activity needs to be converge to a launch event of some sort, calling for a Launch Boss. Too many decisions need to be made, too fast, for things to happen through consensus or deliberation, or go through judicial style review processes by lots of people. So somebody just grabs executive authority to drive the launch through. I used to play this role more in the past but these days it is too intense for me unless it is really short, like running a meeting.

High-stakes, low-energy BDFxing is needed when a collective effort has created a lot of great raw materials that just need a bit of creative catalysis to come together. Like a grain of dust being dropped into a superheated or supercooled liquid. I like the metaphor of a proto Frankenstein monster laying on the operating table, a bunch of parts that have been assembled, but aren’t coming to life. You need a lightning bolt of creative synthesis, but usually not in terms of content. You need a creative frame provided by a Lightning Conductor. I often play this role. Often all a mass of creative content needs to come together is just a 2×2 to structure or a spreadsheet it for divide-and-conquer finishing drive, and I’ve often been the one to supply the 2×2. Lightning conducting looks very similar to traditional project management, but is like doing only the fun creative part, because everyone manages their own work within the creative frame.

Low-stakes, low-energy BDFxing makes for the longest burst of BDFxing. At the Yak Collective, we sometimes refer to this as “tours of duty,” where entropy has accumulated somewhere for a while (like years), but it’s in a state where a few weeks of steady, low-intensity TLC will restore it to full vigor. It calls for a Landscaper. I personally think this is the most important kind of BDFxing, and the hardest to culturally incentivize. Lots of people are willing to handle intense 1-day launch events or less intense 1-week lightning conductor jobs. But one month of housekeeping at a few minutes to an hour a day is a pattern of entropy-arresting energy injection that is really rare, and correspondingly valuable. It is doubly valuable because when something is seen to be done in a disciplined way for a few weeks, other people form good habits around it, which persists even after the period of landscaper BDFxing ends. This is why we also often say about participating in the Yak Collective: 1 hour a week over 10 weeks is more valuable than 10 hours in 1 day.

Low-stakes, high-energy BDFxing is usually called for when something needs a marketing or PR “face.” Unlike traditional leadership where that is a high stakes performance in its own right, the BDFxing version is much lower stakes, since much less actually depends on the leader’s talents, but still calls for high energy. I like the archetype of Ringleader for this. In a circus, the ringleader has to keep the energy high and the energy of the show going, but it’s the acrobats, jugglers and other individuals who are responsible for the stakes. I often play this role as well, and have taken to calling myself a “frontman” when I do. By virtue of being a relatively better known mediocre influencer, in groups with less known people, I often attract more attention than I “deserve” for whatever is happening. Which means the job is really to direct it to the right places.

Everybody should try their hand at all four kinds of BDFxing, and figure out what they’re best at. And then do it for the things they are involved in to the degree it is fun. If nobody has enough fun doing the BDFxing to supply the leadership of the activity, the activity should probably just be abandoned. A leadership deficit that’s fixed by coercive force and misery poisons the activity and the outcome.

Cozy Hypertext for the Dark Forest Web

February 17, 2024 By

We need to reinvent hyperlinks in a cozy idiom for the contemporary dark forest online environment before web2 platforms succeed in their decade long quest to kill them (though they don’t deserve all the blame for hyperlinks being increasingly suboptimal constructs). I trace the origins of this effort to Google’s Amp product which had “faster mobile user experience” as the casus belli for the disproportionately belligerent war on links, which is now starting to look like the Drug War in the US. Now Google routinely hijacks links for shitty and unnecessary eager highlighting and PDF handling. The view of hypertext culture I shared 15 years ago in The Rhetoric of the Hyperlink, which was quite popular at the time, now seems hopelessly idealistic. When I just googled for that link, ironically, I had to strip out some shitty highlighting crud. The war extends even to cozy internal linking within a blog to the extent you rely on external tools, such as by using Google as your internal search tool. Which is why cozy products like Roam and Notion wisely choose to build seamless internal search-and-link-while-writing author experiences (AX) that can’t be “improved” by Google.

The disingenuous philosophy in support of this war is the idea that URLs are somehow dangerous and ugly glimpses of a naked, bare-metal protocol that innocent users must be paternalistically protected from by benevolent and beautiful products. The truth is, when you hide or compromise the naked hyperlink, you expropriate power and agency from a thriving commons. Sure, aging grandpas may have some trouble with the concept but that’s true of everything, including the friendliest geriatric experiences (GXes). My grandfather handled phone numbers and zip codes fine. URLs aren’t much more demanding and vastly more empowering to be able to manipulate directly as a user. Similarly, accessibility considerations are a disingenuous excuse for a war on hyperlinks.

A useful way to think about this is the interaction of the Hypertext Experience (HX) with Josh Stark’s notion of a Trust Experience (TX), which needs to be extended beyond the high-financial-stakes blockchain context he focuses on, to low-stakes everyday browsing. We all agree that the TX of the web has broken and it’s now a Dark Forest. The median random link click now takes you to danger, not serendipitous discovery. This is not entirely the fault of platform corps. We all contributed. And there really is a world of scammers, trolls, phishers, spammers, spies, stalkers, and thieves out there. I’m not proposing to civilize the Dark Forest so we don’t need to protect ourselves from it. I merely don’t want the protection solution to be worse than the problem. Or worse, end up in a “you now have two problems” situation where the HX is degraded with no security benefits, or even degraded security.

[Read more…]

The Dark Forest Marketing Agency

February 15, 2024 By
This entry is part 10 of 10 in the series Fiction

The Pasha sat gravely on a beautifully rendered Ottoman-era chair. A discreet timer hovered above his head, reading 0:15. To one side, the medieval Istanbul cityscape was visible through a window. The Pasha gestured me towards a chair identical to his own. I sat down, briefly registering the dissonance between the visual of the luxuriously cushioned antique chair and the hard barstool I was actually sitting on in my apartment. I did not have the budget for luxuries like adaptive texture chairs and 2d treadmills. But as usual, the dissonance was fleeting.

The Pasha spoke.

”Welcome to the Dark Forest Marketing Agency. I am required by Turkish law to disclose that I am Pasha7, a certified digital twin representing Mr. Ibrahim Pasha of Sultanahmet, the renowned human marketing expert with whom you have signed the contract for this initial fifteen-minute consultation at the low introductory price of 0.05 ETH. As per Turkish law, Mr. Pasha accepts full legal responsibility for my actions on his behalf, as an AI model trained to track and mirror his evolving expertise as a marketing consultant, and offer affordable AI-based counsel on his behalf. I will strive to provide services comparable to the full-price ones provided by Mr. Pasha himself, to the select clients he handles personally. Please indicate your awareness that you are speaking with an AI mirror model.”

[Read more…]

Mediocratopia: 13

February 14, 2024 By
This entry is part 13 of 13 in the series Mediocratopia

Meant to blog this earlier but forgot. A couple of months ago, Eric Platon shared an image of a couple of French books (not available in translation) on mediocrity with this comment:

Just discovering that the meaning of mediocrity changed with Renaissance in Europe (don’t know which one yet). Mediocrity was apparently not pejorative before!…The back cover is rich: Mediocrity used to mean anything “median”, finding compromise , perplexity, mood balance, notion of androgyny, intermediary positioning in politics and religion. Very different from post-Renaissance apparently.

This tracks for me. Mediocrity used to mean something closer to the Buddhist middle path doctrine. It got recoded somehow to mean low quality, apathy, sloppiness and so on.

The point about being on the middle of various spectra rather than at an extreme is that it forces you to acknowledge it as a spectrum. Being at 0 or 1 on an (normalized) parameter allows you to essentially drop a dimension and build in a degeneracy. Being at 0.534 forces you to model the dimension thoughtfully. I’ve said before that mediocrity is about fatness — reserve resources. Now we can add — it’s also about fullness. Full rank. Max dimensionality. No unnecessary reduction to binaries or absolutes. Everything in the nature of the thing is also in the tradeoff space of dealing with the thing. This also means there is something tentative about a mediocre thing. All options are open. None have been foreclosed.

“Excellence” is also about optimization, and to the extent optima occur at boundary extrema rather than interior points, you get a similar degeneracy from the mere fact of using optimization frames. If you’re on a boundary you don’t need to model the interior. This is why mediocritization is the opposite of optimization. The Taoist fable, Maybe so, maybe not, we’ll see captures the spirit of this idea.

The drive to excellence is often a pathological drive towards degeneracy and legibility via a leaning-out of fatness and fullness. And false determinacy. Converting maybes into yeses and noes by fiat, and not waiting to see. It is high modernism — an aesthetic pretending to be an ethic.

I want to connect up one more thought I might elaborate on later. Mediocrity is often more thorough in its consideration of things in all their fullness. Excellence often has a focus on some notion of efficiency. Every notion of optimality is also a notion of efficiency that sacrifices thoroughness to some degree. By the efficiency thoroughness tradeoff (ETTO) principle, you can’t have both at once.

Thoroughness is often associated with craft where efficiency is associated with industry. Craftsmanship does not manifest excellence in the sense a quality industrial product can. It is a not-even-wrong standard to apply. But a well-crafted thing often expresses the fullest nature of a thing. A beautiful wabi-sabi bowl may look rough butcaptures the nature of the bowl, including its transience and mortality, in the fullest way. But it seems strange to think of it as excellent.

BDFxing, Or Post-Charismatic Distributed Leadership

February 7, 2024 By
This entry is part 1 of 2 in the series Lunchtime Leadership

The management cultures I inhabit in my very-online blogger life tend to run a generation ahead of the ones I support in my very-offline consultant life, since I mostly support executives roughly my age (49) or older in traditional orgs. But sometimes, it is helpful to signal-boost management patterns pioneered by younger people, not just because they work better than old patterns in new media organizational contexts (Slack-based orgs for example), but because they work better, period.

One such pattern I strongly recommend you understand and cultivate in your org if you don’t already is the BDFx, or Benevolent Dictator for x, pattern, where x is a time period between an hour or a year or so. The limits vary by context. In various orgs I’m in, it tends to be days to months.

BDFx as a prescriptive term derives from BDFL, L for “Life,” as a descriptive term. That term is old and dates to 1995. It originated in the Python community and is now generally used to describe the condition of an open-source leader who may find themselves saddled with more, and longer-term, expectations of selfless (bordering on martyrdom) leadership than they may want. The condition of the BDFL is captured by this famous xkcd cartoon.

You see, actual leadership is a thankless job even when you’re motivated by, and being rewarded with, great wealth (stock etc), power, and fame (being US President, a Hollywood producer, etc). As I’ve argued before (in a 2015 post) most leaders motivated by those things don’t actually lead. Instead they indulge in a theatrical grifter activity I call leadering, which delivers the rewards without requiring them to deal with the responsibilities. In the open-source world, since these adverse selection incentives are mostly missing, you see more actual leadership, but also more honest appraisals of what leadership is. And more willingness on the part of people who do it to say fuck you and walk away when the job goes from being merely thankless to attracting things worse than ingratitude, like resentment and unfair blame. If you’re toiling away in thankless obscurity with no wealth, fame, or power anyway, the only reason to accept martyrdom is if you’re either a masochist or a true saint defending the world against serious pain. Which is sometimes the case but far rarer than it might seem, since fake saints are ubiquitous.

Now what’s the alternative to selfless leaders being burdened beyond endurance, to the point they say fuck you and walk, causing things to collapse on the rest of us?

[Read more…]