• Machines
• Organisms
• Brains
• Cultures
• Political systems
• Psychic prisons
• Flux and transformation
• Instruments of domination

Whilst reading Helen Czerski’s Blue Machine, it struck me that we need a similar canonical metaphor list for protocols.

Some candidates:

• Oceans
• Musculoskeletal systems
• Civil sociotechnical infrastructure
• Mesoscale biological signalling

• Machines
• Organisms
• Brains
• Cultures
• Political systems
• Psychic prisons
• Flux and transformation
• Instruments of domination

Whilst reading Helen Czerski’s Blue Machine, it struck me that we need a similar canonical metaphor list for protocols.

Some candidates:

• Oceans
• Musculoskeletal systems
• Civil sociotechnical infrastructure
• Mesoscale biological signalling

There are two key emotions that drive my research habits: wonder, and pure unadulterated rage

The Rage of Research - ↗

AI must be a liberating equalizer for humanity. Capable, reliable assistance should reach everyone with a phone or computer, giving them greater command of their time, information and choices. As we prepare for a world with super intelligence, sovereignty must extend to ordinary people’s lives.

Own your AI, don’t let it own you — Underdog ↗

And the shoulders that Steve Jobs, Bill Gates, and others stood on were not those of pocket-protector-wearing nerds clocking into jobs at Lockheed, but young and freaky computer scientists. My friend John Markoff’s book What the Dormouse Said is all about how the counterculture set the tone for the Valley during its tremendous growth years. When we spoke this week, Markoff noted that the Valley’s transition point from a defense-oriented industry to one shaped by counterculture values rose from three pivotal computer labs—AI pioneer John McCarthy’s Stanford lab, the Stanford Research Institute, which hosted Douglas Engelbart’s groundbreaking GUI innovations; and Kay’s Xerox PARC, which essentially set the foundation for the computers that we all use today.

War Bros Didn’t Always Rule Silicon Valley | WIRED ↗

i basically dont have time to write anymore. maybe ill release something once every 1-2 months, and even that is a stretch. im now squarely incapable of doing anything near my prior 1-2 week/article cadence. it was a good 2.5 year run!

owl on X: i basically dont have time to write anymore. / X ↗

First, traditional product thinking is outcome-based. One picks a target state, evaluates risks, and constructs archetypical cases to steer toward it. Success is arriving at the intended destination.

Protocol thinking optimises for something else entirely: an expansion of the moves available to players of the game. The protocol designer does have near-term goals, such as interoperability, legibility, and a low cost of compliance, but beyond them, they deliberately under-determine what gets built. That restraint is the point. Every outcome specified is a move taken off the board for someone else.

This is why indexing on outcomes, or even a cluster of them, fails. The upside a good protocol unlocks isn’t just hard to predict; it’s unnameable in advance. Steering toward a target state, however carefully chosen, narrows the space that’s meant to be widening. The question is whether others can now do surprising things in surprising ways neither they nor we could have specified beforehand.

Second, because protocol thinking solves for, essentially, unanticipated upside and future ecosystem optionality, one’s stance toward value capture and accumulated credit must change dramatically. Not softened. Abandoned.

Credit requires attribution, and attribution requires a nameable outcome. The moment one starts keeping score, one is pulled back toward the outcomes one can point at, which is to say back toward product thinking.

Value capture does the same damage more directly: every toll taken is a tax on the very moves the protocol was meant to make available. The aim is to expand others’ agency, not enlarge the designer’s share.

Protocol thinking in this sense asks for something product thinking never does: surrendering one’s own agency, credit accumulation and value capture in exchange for others’ capacity to accrue them.

Third, inbound changes must be guarded far more judiciously. In product, a feature request is a signal to weigh against a roadmap. In a protocol, a proposed change is an amendment to the rules everyone plays by, and it deserves scrutiny proportional to that, especially when it foregrounds someone’s value capture.

Some open-source projects have closed themselves to outside contributions rather than drown in evaluating them. That’s one answer. The better one is explicit criteria and quality gates that make the bar legible before anything is proposed. The bar must be high because, in a protocol, the letter and the spirit have to be synonymous.

Product teams can lean on intent, norms, and a terms-of-service page. Protocols can’t. Anything the rules permit will eventually be done; anything extractable will be extracted, as maximal extractable value shows with grim reliability. A change that’s harmless in spirit but exploitable in letter is simply exploitable. If the spirit isn’t encoded in the letter, it doesn’t exist.

First, traditional product thinking is outcome-based. One picks a target state, evaluates risks, and constructs archetypical cases to steer toward it. Success is arriving at the intended destination.

Protocol thinking optimises for something else entirely: an expansion of the moves available to players of the game. The protocol designer does have near-term goals, such as interoperability, legibility, and a low cost of compliance, but beyond them, they deliberately under-determine what gets built. That restraint is the point. Every outcome specified is a move taken off the board for someone else.

This is why indexing on outcomes, or even a cluster of them, fails. The upside a good protocol unlocks isn’t just hard to predict; it’s unnameable in advance. Steering toward a target state, however carefully chosen, narrows the space that’s meant to be widening. The question is whether others can now do surprising things in surprising ways neither they nor we could have specified beforehand.

Second, because protocol thinking solves for, essentially, unanticipated upside and future ecosystem optionality, one’s stance toward value capture and accumulated credit must change dramatically. Not softened. Abandoned.

Credit requires attribution, and attribution requires a nameable outcome. The moment one starts keeping score, one is pulled back toward the outcomes one can point at, which is to say back toward product thinking.

Value capture does the same damage more directly: every toll taken is a tax on the very moves the protocol was meant to make available. The aim is to expand others’ agency, not enlarge the designer’s share.

Protocol thinking in this sense asks for something product thinking never does: surrendering one’s own agency, credit accumulation and value capture in exchange for others’ capacity to accrue them.

Third, inbound changes must be guarded far more judiciously. In product, a feature request is a signal to weigh against a roadmap. In a protocol, a proposed change is an amendment to the rules everyone plays by, and it deserves scrutiny proportional to that, especially when it foregrounds someone’s value capture.

Some open-source projects have closed themselves to outside contributions rather than drown in evaluating them. That’s one answer. The better one is explicit criteria and quality gates that make the bar legible before anything is proposed. The bar must be high because, in a protocol, the letter and the spirit have to be synonymous.

Product teams can lean on intent, norms, and a terms-of-service page. Protocols can’t. Anything the rules permit will eventually be done; anything extractable will be extracted, as maximal extractable value shows with grim reliability. A change that’s harmless in spirit but exploitable in letter is simply exploitable. If the spirit isn’t encoded in the letter, it doesn’t exist.

Consider, for instance, the rise of molecular biology as a discipline. We think of Watson and Crick as molecular biologists, not as an ornithologist and a physicist. The first molecular biologists were a motley crew of misfits and revolutionaries with no particularly relevant training, many of them ex-physicists. These physicists didn't waste much time identifying themselves as physicists any more. They viewed themselves as a new kind of biologist. They burned their bridges. Max Delbrück dropped physics and started studying phage replication because it seemed like the fastest, best way to crack the molecular basis of heredity. It's hard to imagine molecular biology making such dramatic progress if it had involved forming interdisciplinary teams of physicists and biologists. The molecular biologists were viewed as naive infidels.

“Antedisciplinary” Science | PLOS Computational Biology ↗

The first thing I have to remind myself is that Product Brain must evolve to Protocol Brain. Don’t make the client or the platform; make the protocol-ish engine that can power numerous arbitrary, value-adding manifestations of clients and platforms. To do that, I’ll take an Amazonian approach and work backwards.

In Hesse’s novel, a Glass Bead Game is played by Castalia’s scholars using a symbolic language that links ideas from every discipline, such as a musical theme, a mathematical proof, or a line of poetry. A player opens with a motif, then develops and transforms it through variations, counterpoints and correspondences across fields, much like composing a fugue.

The aim isn’t to win but to create a game of maximum harmony, elegance and insight, revealing the hidden unity of all knowledge. A great game is judged as a work of art and is also experienced as contemplation, close to prayer.

Happily, Blygger’s ethos and resulting protocol—blogs × wikis × Twitter × git changelogs, built on static files and RSS—provide the primitives to bring GBGs to life on the web, today.

Let’s pretend that one of The Protocol Institute’s Special Interest Groups (SIGs) is extending their roster of events. On a weekly basis, they will be hosting a Glass Bead Game.

The Game functions as follows. The two players involved agree to an exchange which the SIG has a particular interest in. Let’s say the SIG provides a prompt or brief in advance.

The Game begins. The first turn is a Player making an opening move using their own Blyg. Perhaps they iterate through a fragment or piece together a thread. When time runs out, the move is locked in.

The other Player responds via their own Blyg. They may respond directly with a fragment. They might transclude a thread from the Blyg of someone else in the SIG (a number of whom are watching the Game unfold live, hanging out on Discord).

The exchanges go back and forth, with fragments, threads, transclusions, stubs, and off-Blyg excerpts all woven into a dense tapestry of live thought.

At the close of the Game, the entire structure is itself locked up, and then becomes a part of the cultural mosaic, and informs both ongoing Blygging and future Games.

Extending Blygger for GBGs

Working backwards from the Game, the protocol needs to supply a handful of things:

• A brief that convenes the Game and that every move can point back to
• Moves, authored by each player on their own Blyg
• Lineage, so each move names what it builds on
• Lock-in, so a finished move becomes immutable and citable
• Notification, so a player knows the other has moved
• Membership, so any item can say which Game it belongs to
• Outside material, so moves can draw in other Blygs and off-Blyg sources
• A record, so the finished Game becomes a single artefact

Almost all of this already exists.

• The brief is just a fragment or thread on the SIG’s Blyg
• Moves are fragments and threads
• Lineage is stub_of
• Lock-in is a pin, which Blygger already treats as a permanent, irrevocable promise
• Notification is Webmention
• Outside material arrives through cross-origin transclusion, or through a stub of a plain URL
• The record is a thread on the SIG’s Blyg that transcludes every pinned move and is then pinned itself

Timing, turn order, spectators and judging stay out of the protocol entirely, just as AI does.

Only two things are missing:

Membership. A new optional field, game_of, names the brief. Clients that don’t understand it ignore it, so a move still reads as an ordinary stub everywhere else.

Multiple parents. Synthesis moves build on more than one prior move, but stub_of names a single target. Relaxing it to accept several means one Webmention per parent. A more conservative route leaves stub_of alone and treats a move’s transclusions as its extra parents.

A valid move, then, is a pinned L2 item that carries game_of and stubs at least one prior move in the same Game; an opening move stubs the brief. The pinned version is the move, so later edits to the live item don’t touch the Game. The protocol only defines what counts. The SIG’s Game client checks who played and when, and leaves anything invalid out of the record.

A few questions remain open.

• Should references be able to name a specific pinned version, so lineage points at the move rather than the living item?
• Should the brief list the players, so any client can validate a Game?
• Is game_of really a general “convened by” field that salons and reading groups could reuse?

In the meantime, we’ll have to make do with material like this: “We live in a wild era. I bought a smart mattress that tracks my sleep cycles, and now it sends weekly performance reports to my phone. Nothing ruins your Monday morning quite like your bed giving you a C-minus in resting.”

The first thing I have to remind myself is that Product Brain must evolve to Protocol Brain. Don’t make the client or the platform; make the protocol-ish engine that can power numerous arbitrary, value-adding manifestations of clients and platforms. To do that, I’ll take an Amazonian approach and work backwards.

In Hesse’s novel, a Glass Bead Game is played by Castalia’s scholars using a symbolic language that links ideas from every discipline, such as a musical theme, a mathematical proof, or a line of poetry. A player opens with a motif, then develops and transforms it through variations, counterpoints and correspondences across fields, much like composing a fugue.

The aim isn’t to win but to create a game of maximum harmony, elegance and insight, revealing the hidden unity of all knowledge. A great game is judged as a work of art and is also experienced as contemplation, close to prayer.

Happily, Blygger’s ethos and resulting protocol—blogs × wikis × Twitter × git changelogs, built on static files and RSS—provide the primitives to bring GBGs to life on the web, today.

Let’s pretend that one of The Protocol Institute’s Special Interest Groups (SIGs) is extending their roster of events. On a weekly basis, they will be hosting a Glass Bead Game.

The Game functions as follows. The two players involved agree to an exchange which the SIG has a particular interest in. Let’s say the SIG provides a prompt or brief in advance.

The Game begins. The first turn is a Player making an opening move using their own Blyg. Perhaps they iterate through a fragment or piece together a thread. When time runs out, the move is locked in.

The other Player responds via their own Blyg. They may respond directly with a fragment. They might transclude a thread from the Blyg of someone else in the SIG (a number of whom are watching the Game unfold live, hanging out on Discord).

The exchanges go back and forth, with fragments, threads, transclusions, stubs, and off-Blyg excerpts all woven into a dense tapestry of live thought.

At the close of the Game, the entire structure is itself locked up, and then becomes a part of the cultural mosaic, and informs both ongoing Blygging and future Games.

Extending Blygger for GBGs

Working backwards from the Game, the protocol needs to supply a handful of things:

• A brief that convenes the Game and that every move can point back to
• Moves, authored by each player on their own Blyg
• Lineage, so each move names what it builds on
• Lock-in, so a finished move becomes immutable and citable
• Notification, so a player knows the other has moved
• Membership, so any item can say which Game it belongs to
• Outside material, so moves can draw in other Blygs and off-Blyg sources
• A record, so the finished Game becomes a single artefact

Almost all of this already exists.

• The brief is just a fragment or thread on the SIG’s Blyg
• Moves are fragments and threads
• Lineage is stub_of
• Lock-in is a pin, which Blygger already treats as a permanent, irrevocable promise
• Notification is Webmention
• Outside material arrives through cross-origin transclusion, or through a stub of a plain URL
• The record is a thread on the SIG’s Blyg that transcludes every pinned move and is then pinned itself

Timing, turn order, spectators and judging stay out of the protocol entirely, just as AI does.

Only two things are missing:

Membership. A new optional field, game_of, names the brief. Clients that don’t understand it ignore it, so a move still reads as an ordinary stub everywhere else.

Multiple parents. Synthesis moves build on more than one prior move, but stub_of names a single target. Relaxing it to accept several means one Webmention per parent. A more conservative route leaves stub_of alone and treats a move’s transclusions as its extra parents.

A valid move, then, is a pinned L2 item that carries game_of and stubs at least one prior move in the same Game; an opening move stubs the brief. The pinned version is the move, so later edits to the live item don’t touch the Game. The protocol only defines what counts. The SIG’s Game client checks who played and when, and leaves anything invalid out of the record.

A few questions remain open.

• Should references be able to name a specific pinned version, so lineage points at the move rather than the living item?
• Should the brief list the players, so any client can validate a Game?
• Is game_of really a general “convened by” field that salons and reading groups could reuse?

The first thing I have to remind myself is that Product Brain must evolve to Protocol Brain. Don’t make the client or the platform; make the protocol-ish engine that can power numerous arbitrary, value-adding manifestations of clients and platforms. To do that, I’ll take an Amazonian approach and work backwards.

In Hesse’s novel, a Glass Bead Game is played by Castalia’s scholars using a symbolic language that links ideas from every discipline, such as a musical theme, a mathematical proof, or a line of poetry. A player opens with a motif, then develops and transforms it through variations, counterpoints and correspondences across fields, much like composing a fugue.

The aim isn’t to win but to create a game of maximum harmony, elegance and insight, revealing the hidden unity of all knowledge. A great game is judged as a work of art and is also experienced as contemplation, close to prayer.

Happily, Blygger’s ethos and resulting protocol—blogs × wikis × Twitter × git changelogs, built on static files and RSS—provide the primitives to bring GBGs to life on the web, today.

Let’s pretend that one of The Protocol Institute’s Special Interest Groups (SIGs) is extending their roster of events. On a weekly basis, they will be hosting a Glass Bead Game.

The Game functions as follows. The two players involved agree to an exchange which the SIG has a particular interest in. Let’s say the SIG provides a prompt or brief in advance.

The Game begins. The first turn is a Player making an opening move using their own Blyg. Perhaps they iterate through a fragment or piece together a thread. When time runs out, the move is locked in.

The other Player responds via their own Blyg. They may respond directly with a fragment. They might transclude a thread from the Blyg of someone else in the SIG (a number of whom are watching the Game unfold live, hanging out on Discord).

The exchanges go back and forth, with fragments, threads, transclusions, stubs, and off-Blyg excerpts all woven into a dense tapestry of live thought.

At the close of the Game, the entire structure is itself locked up, and then becomes a part of the cultural mosaic, and informs both ongoing Blygging and future Games.

Extending Blygger for GBGs

Working backwards from the Game, the protocol needs to supply a handful of things:

• A brief that convenes the Game and that every move can point back to
• Moves, authored by each player on their own Blyg
• Lineage, so each move names what it builds on
• Lock-in, so a finished move becomes immutable and citable
• Notification, so a player knows the other has moved
• Membership, so any item can say which Game it belongs to
• Outside material, so moves can draw in other Blygs and off-Blyg sources
• A record, so the finished Game becomes a single artefact

Almost all of this already exists.

• The brief is just a fragment or thread on the SIG’s Blyg
• Moves are fragments and threads
• Lineage is stub_of
• Lock-in is a pin, which Blygger already treats as a permanent, irrevocable promise
• Notification is Webmention
• Outside material arrives through cross-origin transclusion, or through a stub of a plain URL
• The record is a thread on the SIG’s Blyg that transcludes every pinned move and is then pinned itself

Timing, turn order, spectators and judging stay out of the protocol entirely, just as AI does.

Only two things are missing:

Membership. A new optional field, game_of, names the brief. Clients that don’t understand it ignore it, so a move still reads as an ordinary stub everywhere else.

Multiple parents. Synthesis moves build on more than one prior move, but stub_of names a single target. Relaxing it to accept several means one Webmention per parent. A more conservative route leaves stub_of alone and treats a move’s transclusions as its extra parents.

A valid move, then, is a pinned L2 item that carries game_of and stubs at least one prior move in the same Game; an opening move stubs the brief. The pinned version is the move, so later edits to the live item don’t touch the Game. The protocol only defines what counts. The SIG’s Game client checks who played and when, and leaves anything invalid out of the record.

A few questions remain open.

• Should references be able to name a specific pinned version, so lineage points at the move rather than the living item?
• Should the brief list the players, so any client can validate a Game?
• Is game_of really a general “convened by” field that salons and reading groups could reuse?

The first thing I have to remind myself is that Product Brain must evolve to Protocol Brain. Don’t make the client or the platform; make the protocol-ish engine that can power numerous arbitrary, value-adding manifestations of clients and platforms. To do that, I’ll take an Amazonian approach and work backwards.

In Hesse’s novel, a Glass Bead Game is played by Castalia’s scholars using a symbolic language that links ideas from every discipline, such as a musical theme, a mathematical proof, or a line of poetry. A player opens with a motif, then develops and transforms it through variations, counterpoints and correspondences across fields, much like composing a fugue.

The aim isn’t to win but to create a game of maximum harmony, elegance and insight, revealing the hidden unity of all knowledge. A great game is judged as a work of art and is also experienced as contemplation, close to prayer.

Happily, Blygger’s ethos and resulting protocol—blogs × wikis × Twitter × git changelogs, built on static files and RSS—provide the primitives to bring GBGs to life on the web, today.

Let’s pretend that one of The Protocol Institute’s Special Interest Groups (SIGs) is extending their roster of events. On a weekly basis, they will be hosting a Glass Bead Game.

The Game functions as follows. The two players involved agree to an exchange which the SIG has a particular interest in. Let’s say the SIG provides a prompt or brief in advance.

The Game begins. The first turn is a Player making an opening move using their own Blyg. Perhaps they iterate through a fragment or piece together a thread. When time runs out, the move is locked in.

The other Player responds via their own Blyg. They may respond directly with a fragment. They might transclude a thread from the Blyg of someone else in the SIG (a number of whom are watching the Game unfold live, hanging out on Discord).

The exchanges go back and forth, with fragments, threads, transclusions, stubs, and off-Blyg excerpts all woven into a dense tapestry of live thought.

At the close of the Game, the entire structure is itself locked up, and then becomes a part of the cultural mosaic, and informs both ongoing Blygging and future Games.

Extending Blygger for GBGs

Working backwards from the Game, the protocol needs to supply a handful of things:

• A brief that convenes the Game and that every move can point back to
• Moves, authored by each player on their own Blyg
• Lineage, so each move names what it builds on
• Lock-in, so a finished move becomes immutable and citable
• Notification, so a player knows the other has moved
• Membership, so any item can say which Game it belongs to
• Outside material, so moves can draw in other Blygs and off-Blyg sources
• A record, so the finished Game becomes a single artefact

Almost all of this already exists.

• The brief is just a fragment or thread on the SIG’s Blyg
• Moves are fragments and threads
• Lineage is stub_of
• Lock-in is a pin, which Blygger already treats as a permanent, irrevocable promise
• Notification is Webmention
• Outside material arrives through cross-origin transclusion, or through a stub of a plain URL
• The record is a thread on the SIG’s Blyg that transcludes every pinned move and is then pinned itself

Timing, turn order, spectators and judging stay out of the protocol entirely, just as AI does.

Only two things are missing:

Membership. A new optional field, game_of, names the brief. Clients that don’t understand it ignore it, so a move still reads as an ordinary stub everywhere else.

Multiple parents. Synthesis moves build on more than one prior move, but stub_of names a single target. Relaxing it to accept several means one Webmention per parent. A more conservative route leaves stub_of alone and treats a move’s transclusions as its extra parents.

A valid move, then, is a pinned L2 item that carries game_of and stubs at least one prior move in the same Game; an opening move stubs the brief. The pinned version is the move, so later edits to the live item don’t touch the Game. The protocol only defines what counts. The SIG’s Game client checks who played and when, and leaves anything invalid out of the record.

A few questions remain open.

• Should references be able to name a specific pinned version, so lineage points at the move rather than the living item?
• Should the brief list the players, so any client can validate a Game?
• Is game_of really a general “convened by” field that salons and reading groups could reuse?

And since, thanks to Enlightenment, citizenship (membership in the political community), nationality, and humanity had been synthetically merged, being expelled from citizenship meant, quite literally, exclusion from humanity. Hence civic death was necessarily followed by natural death, that is, violent death, or death tout court. Fascist or Nazi genocide was not preceded by legal condemnation (not even in the stunted and fraudulent shape of the so-called administrative verdicts of Cheka “tribunals”): it was the “naturalization” of a moral judgment that deemed some types of human condition inferior. And since there was no protection outside citizenship, lack of citizenship had become the cause of the cessation of the necessary precondition of the human condition—life.

On Post-Fascism - Boston Review ↗

• Must be free
• Must be non-productivity
• Must be mobile
• Must optimise default behaviour

My emotions were never “irrational”—they were just locally optimal.

Why we create our suffering - by Chris Lakin ↗

• Must be free
• Must be non-productivity
• Must be mobile
• Must optimise default behaviour

Under either scenario, the ideological position of the homoploutic class as a legitimate meritocracy will collapse. Under entrenchment, the homoploutic claim to merit becomes untenable because there is no path to entry: Once inheritance is the only way into the top decile of capitalists, such a class can’t credibly describe itself as meritocratic, no matter how hard-working. Under dissolution, if education no longer commands a wage premium, there is no longer any meritorious activity to point at — one is simply an owner of capital, as one’s great-grandfather was. In the first case the asset value of the class survives, but the flow into it stops. In the second, both go.

The End Of Upward Mobility ↗

There it is, the rad­ical offering of 21st cen­tury com­puting, here at last: ask for what you want. It takes prac­tice! And, hon­estly, a cer­tain audacity.

'Ask for what you want' is a key skill for the 21st century ↗

In short, model capability is good enough that compelling products — products that actually have moats — can now be built, and from a business perspective it would do Anthropic and OpenAI good to devote more of their resources to actually building such products.

Frontier Overhangs Stratechery by Ben Thompson ↗

People think that the main benefit of writing in the public domain is “exposure”, or that it compels you to think with more clarity/rigour.

Both are true, but both are subservient to the fact that writing in public leads to a handful of unexpected and meaningful relationships.

• Bohr’s quadrant (pure basic research) seeks fundamental understanding with no consideration of use (e.g. Bohr’s atom, general relativity and the Higgs boson)
• Pasteur’s quadrant (use-inspired basic research) seeks fundamental understanding and is driven by consideration of use (e.g. Pasteur’s germ theory and Shannon’s information theory)
• Edison’s quadrant (pure applied research) is driven by consideration of use with no quest for fundamental understanding (e.g. Edison’s filament trials, the Wright brothers’ wind tunnel and most product R&D)
• Feynman’s quadrant (FAFO, infinite gaming, serious play etc.) is not driven by consideration of use and is not concerned with fundamental understanding (e.g. Geim’s Friday night experiments)

What is increasingly interesting to me is the transitions between these quadrants. There are three traditional routes.

First, there’s the case where fundamental discovery leads to use. Quantum mechanics was worked out in the 1920s as pure physics, and two decades later it made the transistor possible.

Second, there’s the case where the realities of application and practice are pacing the frontier and our elucidation of fundamental truths is scrambling to catch up. The steam engine was running for a century before Carnot and his successors worked out the thermodynamics explaining why.

Third, there’s the case where play turns up something one of the other quadrants picks up. Feynman’s idle maths on a wobbling cafeteria plate led him back to the electron spin problems behind his Nobel.

Historically, these routes between quadrants were structurally enforced by the relationships between the forces of capital, labour and technology. Some definitions:

• Capital: distilled, discrete units of liquid, agentic potential as represented by assets and resources with mutual, agreed upon value.
• Technology: the evolving collection of natural phenomena captured within an assemblage of socio-technical systems and practices that allows a civilisation to fulfil diverse purposes.
• Labour: the purposeful and systematic conversion of energy and matter into forms of progressively higher complexity and meaning undertaken by organic agents.

Generally speaking, research moved at the speed of people. Even with Manhattan projects and fast initiatives, people were the constraint. That’s changed.

The schtick is that technology and capital are increasingly aligning to minimise the reliance on labour supply. That’s why there are big bets being made on 1) personal productivity agents and 2) autonomous meta-science and experimentation. Maximised personal productivity plus accelerated learning loops is the holy grail of the permanent overcast.

Of course, it’s not as simple as that and the world will be curiously playable to a greater extent than ever before. That’s because different paths between the quadrants will dominate and the time taken to walk those paths will be slashed to near-zero. For example:

• FAFO to Pasteur and Bohr becomes routine; exploration gets cheap enough to run wastefully, wildly, and in volume, so play reliably turns up real results
• The Bohr to Edison lag collapses into Pasteur; understanding and use emerge together instead of decades apart
• Edison to Bohr becomes a main road; we build systems that work and accelerate before we even understand them, then we retroactively science ‘em up

If we’re truly entering the Liveness machine, we’re going to have to play seriously. Which means that Feynman’s quadrant, the very one Stokes under-appreciated, will matter much more than we ever expected.

Cloud seeding is undertaken by dispersing substances into the air that serve as cloud condensation or ice nuclei. Common agents include silver iodide, potassium iodide, and dry ice, with hygroscopic materials like table salt gaining popularity due to their ability to attract moisture. Techniques vary from static seeding, which encourages ice particle formation in supercooled clouds to increase precipitation, to dynamic seeding, designed to enhance convective cloud development through the release of latent heat.

Cloud seeding - Wikipedia ↗

"Enshittification" is a sort of "safe horny" term for what is really just outright impoverishment and declining living standards due to economic insolvency from deindustrialization, demographic decline, hyperinflation, etc. You are just de-developing and becoming poor.

mmjukic / status / 2105602006719348751 (x.com) - s:12 ↗

Suddenly, everything has changed. Computers now work the way I thought they did when I was 8.

What Even Is An OS Now? ↗

Kim Heon-sik, a professor at Jungwon University, has compared these sites to mukbang culture — an online culture where viewers watch celebrities or social media influencers eat food. He argues that both dopamine-hit sites and mukbang content let consumers experience a hit of dopamine or mental satiation at the press of a button, without the real-world consequences of mindless eating or social media posting and commenting.

Dopamine sites - Wikipedia ↗

When talented young grappler Tay enters his life, Rick reluctantly finds himself drawn back into the sport he thought he’d left behind. Their turbulent coach-student relationship forces Rick to confront his ego, failures and lost sense of purpose, culminating in a high-stakes championship where his unusual ability to read the game may be Tay’s only chance of pulling off an impossible upset.

WHITE BELT TRAILER - YouTube ↗

Worst case, the direct line to the machine god becomes a sycophantic amplifier, entrenching irrational behaviour and segregating everyone. Best case, a higher value placed on truth and wisdom (perhaps simply because their opposites become unfashionable) lets cultural evolution outcompete the provably irrational faster than ever.

Anabolic Note-Taking is any system which skews toward building, rather than analyzing. For example – the way some people use Notes here is highly anabolic. It’s a combination of distilling concepts, poking the world to see what pops out, testing new language, sacrificing bits of your ego to the feed… then synthesizing the pieces into a coherent article. Those articles then build into a broader series or framework. That framework turns into leverage inside of an organization.

Anabolic Note-Taking - by Timber - Blundercheck ↗

I’m now reading about the different Irish words for “field”, and in this text, Manchan explicitly makes the point that Irish cultural heritage should be treated as indigenous, as opposed to a branch of European culture.

In 2024, I travelled around portions of Australia. Naturally, the indigenous heritage of the place is a big deal. So when Manchan pointed out the Irish-Australia indigenous simultaneity, I think it was the first time I really got the notion of place. Like, in my bones.

Originally, it was Aneesh Sathe who put me onto the salience of place and space, and the conceptual underpinnings developed by Yi-Fu Tuan. For example, in his post on Nakatomi Drift, Aneesh writes:

“Yi-Fu Tuan was a philosopher and geographer who wrote, “Place is security, Space is freedom.” For Tuan, Space and Place are not simply large versus small geographic areas. Space is associated with movement, openness, freedom, possibility, and exposure. Place arises through pause, familiarity, value, and orientation. Crucially, each requires the other: movement through Space produces Places, while established Places allow us to perceive and venture into Space.”

This matters because there’s a chance—small, precious, fleeting—that Blygs become places. We may end up with rich, storied equivalents to places like the Connemaran Magairlí an Deamhain, the demon’s testicles.

But this phenomenon is not limited to weekdays. In the full paper, we find analogous structure in language models across more complex settings, including months, letters, ages, and a synthetic in-context learning task with predefined geometries. We also extend our results across modalities with an image-action model predicting the position of a car rolling up and down a hill (see the demo in our previous post). Across these tasks, manifold geometry provides a practical blueprint for steering model behavior.

Steering Along Manifolds to Control Neural Networks - Goodfire ↗

The relevant success metrics for AI in drug discovery need to be numbers that matter, rather than contrived proxy measures that primarily optimize for scientific publications rather than actual R&D progress. Ultimately, what truly matters is increasing clinical approval rates, as well as novelty in both modality and mode of action space.

Artificial intelligence in drug discovery — what it is, where we stand and the path forward | Nature Reviews Drug Discovery ↗

The large volume of data produced by AI agents is posing a problem both for monitoring and analyzing their behavior. While investigating the Hugging Face incident, METR had to analyze immense amounts of data in a very short time. They could only do it by using AI agents themselves and they had to use one of the very same models that were involed in the Hugging Face incident.

Robert Herr (@krherr): The bad news just keeps coming... ↗

We conjecture that the complexity of chocolate implies the possibility of superchocolate. Chocolate didn’t have to exist. If the cacao tree had gone extinct before humans arrived, the particular point in flavor space that chocolate encodes would simply be unknown to us. Given how extraordinarily large flavor space is, it seems likely that other exceptionally delicious embeddings exist that no human has ever experienced. And there is no particular reason to think that chocolate represents a maximum of delight. The right combination of molecules might make chocolate look like oatmeal.

Four Biological Conjectures ↗

We need Copernican frames because most arguments about artificial intelligence today race past foundational ontological questions to ethical ones posed in what are effectively Ptolemaic frames. We ask, within unwieldy anthropocentric frames (with vague and ill-posed adjectives like “super” or “general” serving as epicycles), whether AI will be good or bad, whether it will replace workers, accelerate science, undermine democracy, achieve superintelligence or destroy humanity, all before adequately exploring what actually existing AI even is, and what its actual evolutionary dispositions are.

Our Eukaryotic Moment - by Venkatesh Rao - Contraptions ↗

A working group on the care of older people proposed twelve principles of a good death in their final report. Starting with “To know when death is coming, and to understand what can be expected,” the list closes with “To be able to leave when it is time to go, and not to have life prolonged pointlessly.”

Good Death Web - Summer of Protocols ↗

In Hesse’s novel, The Glass Bead Game is about weaving together disparate ideas into beautiful constructs. A monastic order is built around it. I’m thinking that such a game could use Blygger’s transclusion and pin mechanics to do something similar.

Person A posts a fragment. Person B replies with a fragment that complements or extends or refutes the ideas. Person B responds. This back and forth goes on, in a rapid flurry, and the aim for both is to play a beautiful game of intellectual exchange.

An example:

1 · Maya (opening bead) A fugue states a theme, then states it again over itself, and the second statement changes what the first one meant.

2 · Jonah → builds on 1 Paths across a lawn work the same way. The first walker is a guess; the hundredth is a road.

3 · Maya → builds on 2 Desire lines are votes cast with feet. Cities that pave them end up with the most honest maps.

4 · Jonah → builds on 1 Oral poets didn’t store poems; they re-made them. Each recitation was the poem, reshaped to fit the room.

5 · Maya → builds on 4 So an edit isn’t a betrayal of the original. It’s the next performance.

6 · Jonah → builds on 3 The honest map has a blind spot: it records where people went, never where they wanted to go but couldn’t.

7 · Maya → builds on 2 + 4 (synthesis) A path and a poem are both made by passing through. Neither exists between uses; both stay alive by being travelled.

8 · Jonah → builds on 6 + 7 (synthesis) Then pinning is paving. It stops the path wandering, which saves it and ends it at once.

This happened natively and organically on Twitter where you’d have an exchange between two people kick off. It could be the organising mechanic for Blygger. Everyone brings their fragments and the aim is to play beautiful games with one another.

But it’s the elevation to a key event that differentiates it. It’s not adversarial debate, or passive commentary or sub-tweeting; it’s more like Johnstone-style, carnival-time improv. And it’s the whole point, not a happy accident.

That is why pluralism matters. True pluralism is not adding another term to the utility function for “other perspectives.” It is accepting that other traditions may possess ways of knowing and not-knowing that cannot be losslessly translated into your framework, that conflicting goods may remain genuinely conflicting, and that no community earns jurisdiction over everybody else’s moral reasoning merely by being unusually articulate about its own. That sometimes the right way to deal with visceral doubt is to assign the benefit of the doubt to someone other than yourself, who thinks very differently.

EA Safety - by Venkatesh Rao - Contraptions ↗

In fact, classical economic theory assigns this exact function to entrepreneurs. Frank Knight argued that profits arise not from bearing calculable risk, but from navigating situations where the future cannot be predicted. The entrepreneur’s role is to convert uncertainty about markets, technologies, and customers into functioning businesses.

The Founder–Venture Capital Paradox: Moonshots, Survival, and the Resolution of Uncertainty ↗