{
  "blyg": "0.3",
  "id": "5vt4cvsxvs6h2m5yqzk2hn4d4v",
  "kind": "fragment",
  "version": 12,
  "at": "2026-10-17T20:35:02Z",
  "note": null,
  "pinned": true,
  "origin": "https://www.msweet.net/notes/",
  "author": {
    "name": "Matthew McDowell-Sweet",
    "url": "https://www.msweet.net/"
  },
  "content_md": "**Protocol Thinking for Product People, Or How to Move Slow and Unlock Things**\n\n[Protocol thinking](https://summerofprotocols.com/the-unreasonable-sufficiency-of-protocols-web) is way more distinct from product thinking than I appreciated. Not different by degree. Not even different by kind. It’s different at the [paradigmatic](https://donellameadows.org/archives/leverage-points-places-to-intervene-in-a-system) level. In short, it means:\n\n1. Expanding available moves over achieving specified outcomes\n2. Surrendering majority stakes in value capture and credit\n3. Higher standards and tighter reins around change governance\n4. Evolutionary search and selection over solutions engineering\n5. Care given via individual stewardship and collective goodwill, not exchange\n6. Measuring what’s given away to the wider world, not what’s captured\n7. Making constant failure and certain death generative\n8. De-centring humans amongst many other participating agents\n\n---\n\n**First**, traditional product thinking is [outcome-based](https://www.producttalk.org/product-discovery/). One picks a target state, evaluates [risks](https://swellandcut.com/2025/07/24/the-sixth-product-risk/), and constructs [archetypical cases](https://swellandcut.com/2023/11/08/archetypical-case-making/) to steer toward it. Success is arriving at the intended destination.\n\nProtocol thinking optimises for something else entirely: an expansion of the moves available to [players of the game](https://contraptions.venkateshrao.com/p/the-curiously-playable-universe?r=233ko&triedRedirect=true). 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.\n\nThis 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.\n\n**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. Largely relinquished.\n\nCredit 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.\n\nValue 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.\n\nProtocol thinking in this sense asks for something product thinking never does: exercising one’s agency in setting the rules, then surrendering most claims to credit and value capture in exchange for others’ capacity to accrue them.\n\n**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.\n\nThis isn’t at odds with expanding available moves; it’s what makes it possible. The rules have to be conservative precisely so the play can be permissive. *Move slow and unlock things*.\n\nSome 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.\n\nProduct 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](https://ethereum.org/en/developers/docs/mev/) 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.\n\n**Fourth**, it is more a game of evolution than engineering. Discovery-delivery loops, and even the market-level kinetics [Wardley Maps](https://www.wardleymaps.com/) describe, assume that search is expensive: one reasons one’s way to a good design because one can’t afford to try many. AI breaks that assumption.\n\nWith general intelligence on tap, generating and testing variants becomes cheap enough that search and selection outcompete design, [especially in formalised domains](https://github.com/openai/math). What survives isn’t what was best engineered, but what best satisfies global constraints, fits hyper-local [last-mile requirements](<https://en.wikipedia.org/wiki/Last_mile_(transportation)>), and suits the tastes of the particular agents doing the selecting.\n\nThe protocol’s job is to be a good substrate for that search, not to anticipate its results. It’s akin to an [immune system](https://en.wikipedia.org/wiki/Immune_system) with its valence inverted: both generate variants no one specified and let selection sort them, but one does so to eliminate the novel, the other to propagate it.\n\nIn a sense, protocol thinking is the [Bitter Lesson incarnate](https://msweet.net/notes/117-fern-terns) for product people.\n\n**Fifth**, because of points one to four, protocols don’t fit the constraints of market logic. Transactional exchange, the canonical *quid pro quo*, powers projects, platforms, products and services, but not protocols. Every funding mechanism one reaches for asks for something a protocol can’t give without ceasing to be one.\n\n- State funding needs outcomes it can justify to voters\n- Standards bodies answer to members who eventually steer toward their own capture\n- Community groups have the spirit but not the means, so care becomes unpaid heroics\n- Academia runs on attribution, not on maintaining systems\n- Venture capital needs an order-of-magnitude exit multiple and a capturable share\n- Crypto builds value capture into the rules themselves, which turns governance into a contest over the spoils\n\nNone of these work. That’s because each is a form of exchange, and protocols belong to a different paradigm: the *gift*.\n\nExchange settles accounts so both parties can walk away. A gift keeps moving and creates a loose, unenforceable obligation to pass something on. A protocol gives moves away and asks only that players keep the game going—ideally acting to make the game [more interesting](https://en.wikipedia.org/wiki/Finite_and_Infinite_Games) for others.\n\nGift economies are good at circulation but bad at upkeep. Nobody owes the giver anything, so nobody is on the hook for the roof. Protocol maintenance today rests on collective goodwill and on individuals who take up a role as [benevolent dictators for a bounded period](https://ribbonfarm.com/series/lunchtime-leadership/), carrying the cost of care and prophetic cognition for a while before handing it on.\n\nThe bound matters, because without it such stewardship hardens into ownership. Neither collective goodwill nor individual stewardship is really a funding model—at least not in the sense of capital compensating labour for technological development. How protocols pay for their own care, sadly, remains unsolved.\n\n**Sixth**, in product thinking, most things one cares to measure are proxies for what the market legitimises as value. CAC, LTV, retention, conversion: each stands in for the same question. Namely: *how much of the value generated does the generator get to keep?*\n\nProtocols, as established in point five, sit outside market logic. Their metrics *can’t be* proxies for capture, because capture is the very thing they forgo. Yet they still need *something* to be proxies for.\n\nThat something is value generated but not captured. It’s what a protocol is for, and it’s almost impossible to measure directly. No transaction happens. No one invoices. Nobody is counting. That’s the point. So, as with CAC and LTV, one does need proxies. Four will do:\n\n- **Different people**: the game is played by many players and owned by none\n- **Different implementations**: the spec, not a codebase, holds things together\n- **Different uses**: the play has spread to surprises beyond original intent\n- **Durability**: all three persist over time, by choice, because exit is easy\n\nAny one of these is cheap to fake. All four at once, sustained, is not.\n\n- Fake adopters share an implementation\n- Fake implementations serve a single use\n- Fake uses come from the same few hands\n- Durability turns every fake into a running cost with no market return\n\nThey’re a set, not a menu. Game them for long enough and the four dimensions become indistinguishable from the real thing: a successful, sustainable protocol.\n\nThere are two things they deliberately don’t do, though. They don’t judge whether the play is good; that’s what the rules are for, per point three. And they don’t forecast. The four dimensions are organic, real-time signals, not targets one steers toward, because selection does the steering, per point four. The moment they become targets, for funders or anyone else, the paradigm of exchange dominates again and we bow out of protocol land.\n\nAssuming the market can’t be the measure for a protocol, perhaps the M in TAM, SAM and SOM can stand for something else? Meaning added to the universe.\n\n**Seventh**, product thinking treats failure as an event and death as a decision. Protocols get neither.\n\nComplex systems [always run partly broken](https://how.complexsystems.fail/), and “root cause” is mostly a social need for blame. Blame, like credit, needs attribution, which protocols have already given up. A protocol fails constantly, in small ways, and is held together by whoever’s there to keep catching it.\n\nThe danger comes when no one’s around. Disasters [rarely come out of nowhere](https://hdsr.mitpress.mit.edu/pub/tarzjxgu/release/2): the risks were known, maintenance was undervalued, and the system drifted. Death is [a decision as much as an event](https://summerofprotocols.com/good-death-web), and most protocols have no one with the authority to make it. So they linger as zombies, held up by lock-in and inertia, with live rules and absent stewards. Anything the rules permit will eventually be done, so a dead steward and live rules is open season.\n\nManaged death means planning the ending while the protocol is still healthy: sunset criteria, succession, deprecation timelines. It also means leaving something behind. [Booms](https://press.stripe.com/boom) burn fortunes but leave track others can run on. A protocol that dies well does the same, leaving its spec, namespaces and habits as ground for what comes next. Move slow, unlock things, and when the time comes, die generously.\n\n**Eighth**, in product thinking, not all agents are first-class. As a general rule, product thinking treats human agents as superior to machine ones. Even in contemporary situations, such as agent infra startups, human agents are, at worst, first among equals.\n\nThis primacy made sense when humans were probabilistic beasts and machines were deterministic engines. Humans decided what to do and how to say it, and machines did exactly as they were told. That narrowed the possibility space to four outcomes:\n\n- **Right thing, right way**: the human chose well and specified well, and the machine delivered exactly that\n- **Right thing, wrong way**: the intent was sound but the instruction wasn’t, and the machine faithfully executed the bug\n- **Wrong thing, right way**: the instruction was precise but the intent was misguided, and the machine perfected the mistake at scale\n- **Wrong thing, wrong way**: both failed, usually visibly enough to be caught first\n\nEvery outcome traces back to a human. This was a consequence of computers, for better or worse, doing precisely what they’re instructed to do. It was also a consequence of humans, despite being ever more collectively corralled by sociocultural systems, creatively subverting instructions. The machine contributed nothing but fidelity, so the human was the only place judgement, and therefore error, could live.\n\nNowadays, a whole new class of machines isn’t so deterministic, and societal mechanisms of control are fracturing and fraying in step. Probabilistic machines reopen the possibility space with two outcomes the old one ruled out:\n\n- **Human wrong, machine right**: the machine catches the human’s mistake and does what was meant rather than what was said\n- **Human right, machine wrong**: the machine drifts, reinterprets or confabulates, and the error starts with it rather than the human\n\nOnce machines can subvert instructions too, outcomes no longer trace back to a single human. Judgement and error are distributed, and attribution, the thing credit and blame both depend on, starts to dissolve.\n\nIt gets more complicated, though, because the action is migrating to scales humans can’t inhabit.\n\nAt one end, high-frequency trading and agent swarms create miniaturised economies too fast and too small for humans to make sense of. At the other, exocapitalism is subliming capital from reality. Meanwhile, we’re having to work out Long Now problems we’ve been putting off for generations, problems too slow and too large for any one lifetime.\n\nLived experience has been here before. When steam replaced sail, the sailor’s hard-won feel for wind and weather gave way to stokers shovelling coal below deck. The ship still crossed the ocean and life aboard went on, but the people aboard no longer had a feel for how.\n\nThe upside is that all that complexity can be compressed into protocols. It just takes thought, human or otherwise. The cost is that human lived experience gets well and truly de-centred. It isn’t erased, though. It goes on, as rich as ever, but as one integral thread in a many-dimensional fabric of reality rather than the loom it’s woven on. For as long as we insist on being the centre, the fullest upside of protocols stays locked away.\n\n---\n\nA few critical questions remain in order to truly describe the nature of protocol thinking for product people. A running list for reference:\n\n- The enduring place and value of product thinking\n- The most appropriate [metaphors](https://msweet.net/notes/112-pearl-falcons) for protocol thinking\n- Protocol thinking vs projects, platforms and other “P” paradigms\n- Formal verification and proof interconnection\n- The historical context of protocols for product (e.g. ZIRP era exit, World Machines)\n\nOngoing gratitude to the [Blygger contributors](https://blygger.org/contributors/) for inspiring the above.\n\nAlso of note: soon to be a [book](https://protocol-institute.org/projects/project/?slug=protocol-thinking-for-product-people) as part of the Protocol Institute’s 2026 [PiBoWriMo](https://protocol-institute.org/events/pibowrimo-2026/).",
  "content_html": "<p><strong>Protocol Thinking for Product People, Or How to Move Slow and Unlock Things</strong></p>\n<p><a href=\"https://summerofprotocols.com/the-unreasonable-sufficiency-of-protocols-web\">Protocol thinking</a> is way more distinct from product thinking than I appreciated. Not different by degree. Not even different by kind. It’s different at the <a href=\"https://donellameadows.org/archives/leverage-points-places-to-intervene-in-a-system\">paradigmatic</a> level. In short, it means:</p>\n<ol>\n<li>Expanding available moves over achieving specified outcomes</li>\n<li>Surrendering majority stakes in value capture and credit</li>\n<li>Higher standards and tighter reins around change governance</li>\n<li>Evolutionary search and selection over solutions engineering</li>\n<li>Care given via individual stewardship and collective goodwill, not exchange</li>\n<li>Measuring what’s given away to the wider world, not what’s captured</li>\n<li>Making constant failure and certain death generative</li>\n<li>De-centring humans amongst many other participating agents</li>\n</ol>\n<hr />\n<p><strong>First</strong>, traditional product thinking is <a href=\"https://www.producttalk.org/product-discovery/\">outcome-based</a>. One picks a target state, evaluates <a href=\"https://swellandcut.com/2025/07/24/the-sixth-product-risk/\">risks</a>, and constructs <a href=\"https://swellandcut.com/2023/11/08/archetypical-case-making/\">archetypical cases</a> to steer toward it. Success is arriving at the intended destination.</p>\n<p>Protocol thinking optimises for something else entirely: an expansion of the moves available to <a href=\"https://contraptions.venkateshrao.com/p/the-curiously-playable-universe?r=233ko&amp;triedRedirect=true\">players of the game</a>. The protocol designer <em>does</em> 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.</p>\n<p>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 <em>unnameable in advance</em>. 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.</p>\n<p><strong>Second</strong>, 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. Largely relinquished.</p>\n<p>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.</p>\n<p>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 <em>others’</em> agency, not enlarge the designer’s share.</p>\n<p>Protocol thinking in this sense asks for something product thinking never does: exercising one’s agency in setting the rules, then surrendering most claims to credit and value capture in exchange for others’ capacity to accrue them.</p>\n<p><strong>Third</strong>, 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.</p>\n<p>This isn’t at odds with expanding available moves; it’s what makes it possible. The rules have to be conservative precisely so the play can be permissive. <em>Move slow and unlock things</em>.</p>\n<p>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.</p>\n<p>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 <a href=\"https://ethereum.org/en/developers/docs/mev/\">maximal extractable value</a> shows with grim reliability. A change that’s harmless in spirit but exploitable in letter is simply <em>exploitable</em>. If the spirit isn’t encoded in the letter, it doesn’t exist.</p>\n<p><strong>Fourth</strong>, it is more a game of evolution than engineering. Discovery-delivery loops, and even the market-level kinetics <a href=\"https://www.wardleymaps.com/\">Wardley Maps</a> describe, assume that search is expensive: one reasons one’s way to a good design because one can’t afford to try many. AI breaks that assumption.</p>\n<p>With general intelligence on tap, generating and testing variants becomes cheap enough that search and selection outcompete design, <a href=\"https://github.com/openai/math\">especially in formalised domains</a>. What survives isn’t what was best engineered, but what best satisfies global constraints, fits hyper-local <a href=\"https://en.wikipedia.org/wiki/Last_mile_(transportation)\">last-mile requirements</a>, and suits the tastes of the particular agents doing the selecting.</p>\n<p>The protocol’s job is to be a good substrate for that search, not to anticipate its results. It’s akin to an <a href=\"https://en.wikipedia.org/wiki/Immune_system\">immune system</a> with its valence inverted: both generate variants no one specified and let selection sort them, but one does so to eliminate the novel, the other to propagate it.</p>\n<p>In a sense, protocol thinking is the <a href=\"https://msweet.net/notes/117-fern-terns\">Bitter Lesson incarnate</a> for product people.</p>\n<p><strong>Fifth</strong>, because of points one to four, protocols don’t fit the constraints of market logic. Transactional exchange, the canonical <em>quid pro quo</em>, powers projects, platforms, products and services, but not protocols. Every funding mechanism one reaches for asks for something a protocol can’t give without ceasing to be one.</p>\n<ul>\n<li>State funding needs outcomes it can justify to voters</li>\n<li>Standards bodies answer to members who eventually steer toward their own capture</li>\n<li>Community groups have the spirit but not the means, so care becomes unpaid heroics</li>\n<li>Academia runs on attribution, not on maintaining systems</li>\n<li>Venture capital needs an order-of-magnitude exit multiple and a capturable share</li>\n<li>Crypto builds value capture into the rules themselves, which turns governance into a contest over the spoils</li>\n</ul>\n<p>None of these work. That’s because each is a form of exchange, and protocols belong to a different paradigm: the <em>gift</em>.</p>\n<p>Exchange settles accounts so both parties can walk away. A gift keeps moving and creates a loose, unenforceable obligation to pass something on. A protocol gives moves away and asks only that players keep the game going—ideally acting to make the game <a href=\"https://en.wikipedia.org/wiki/Finite_and_Infinite_Games\">more interesting</a> for others.</p>\n<p>Gift economies are good at circulation but bad at upkeep. Nobody owes the giver anything, so nobody is on the hook for the roof. Protocol maintenance today rests on collective goodwill and on individuals who take up a role as <a href=\"https://ribbonfarm.com/series/lunchtime-leadership/\">benevolent dictators for a bounded period</a>, carrying the cost of care and prophetic cognition for a while before handing it on.</p>\n<p>The bound matters, because without it such stewardship hardens into ownership. Neither collective goodwill nor individual stewardship is really a funding model—at least not in the sense of capital compensating labour for technological development. How protocols pay for their own care, sadly, remains unsolved.</p>\n<p><strong>Sixth</strong>, in product thinking, most things one cares to measure are proxies for what the market legitimises as value. CAC, LTV, retention, conversion: each stands in for the same question. Namely: <em>how much of the value generated does the generator get to keep?</em></p>\n<p>Protocols, as established in point five, sit outside market logic. Their metrics <em>can’t be</em> proxies for capture, because capture is the very thing they forgo. Yet they still need <em>something</em> to be proxies for.</p>\n<p>That something is value generated but not captured. It’s what a protocol is for, and it’s almost impossible to measure directly. No transaction happens. No one invoices. Nobody is counting. That’s the point. So, as with CAC and LTV, one does need proxies. Four will do:</p>\n<ul>\n<li><strong>Different people</strong>: the game is played by many players and owned by none</li>\n<li><strong>Different implementations</strong>: the spec, not a codebase, holds things together</li>\n<li><strong>Different uses</strong>: the play has spread to surprises beyond original intent</li>\n<li><strong>Durability</strong>: all three persist over time, by choice, because exit is easy</li>\n</ul>\n<p>Any one of these is cheap to fake. All four at once, sustained, is not.</p>\n<ul>\n<li>Fake adopters share an implementation</li>\n<li>Fake implementations serve a single use</li>\n<li>Fake uses come from the same few hands</li>\n<li>Durability turns every fake into a running cost with no market return</li>\n</ul>\n<p>They’re a set, not a menu. Game them for long enough and the four dimensions become indistinguishable from the real thing: a successful, sustainable protocol.</p>\n<p>There are two things they deliberately don’t do, though. They don’t judge whether the play is good; that’s what the rules are for, per point three. And they don’t forecast. The four dimensions are organic, real-time signals, not targets one steers toward, because selection does the steering, per point four. The moment they become targets, for funders or anyone else, the paradigm of exchange dominates again and we bow out of protocol land.</p>\n<p>Assuming the market can’t be the measure for a protocol, perhaps the M in TAM, SAM and SOM can stand for something else? Meaning added to the universe.</p>\n<p><strong>Seventh</strong>, product thinking treats failure as an event and death as a decision. Protocols get neither.</p>\n<p>Complex systems <a href=\"https://how.complexsystems.fail/\">always run partly broken</a>, and “root cause” is mostly a social need for blame. Blame, like credit, needs attribution, which protocols have already given up. A protocol fails constantly, in small ways, and is held together by whoever’s there to keep catching it.</p>\n<p>The danger comes when no one’s around. Disasters <a href=\"https://hdsr.mitpress.mit.edu/pub/tarzjxgu/release/2\">rarely come out of nowhere</a>: the risks were known, maintenance was undervalued, and the system drifted. Death is <a href=\"https://summerofprotocols.com/good-death-web\">a decision as much as an event</a>, and most protocols have no one with the authority to make it. So they linger as zombies, held up by lock-in and inertia, with live rules and absent stewards. Anything the rules permit will eventually be done, so a dead steward and live rules is open season.</p>\n<p>Managed death means planning the ending while the protocol is still healthy: sunset criteria, succession, deprecation timelines. It also means leaving something behind. <a href=\"https://press.stripe.com/boom\">Booms</a> burn fortunes but leave track others can run on. A protocol that dies well does the same, leaving its spec, namespaces and habits as ground for what comes next. Move slow, unlock things, and when the time comes, die generously.</p>\n<p><strong>Eighth</strong>, in product thinking, not all agents are first-class. As a general rule, product thinking treats human agents as superior to machine ones. Even in contemporary situations, such as agent infra startups, human agents are, at worst, first among equals.</p>\n<p>This primacy made sense when humans were probabilistic beasts and machines were deterministic engines. Humans decided what to do and how to say it, and machines did exactly as they were told. That narrowed the possibility space to four outcomes:</p>\n<ul>\n<li><strong>Right thing, right way</strong>: the human chose well and specified well, and the machine delivered exactly that</li>\n<li><strong>Right thing, wrong way</strong>: the intent was sound but the instruction wasn’t, and the machine faithfully executed the bug</li>\n<li><strong>Wrong thing, right way</strong>: the instruction was precise but the intent was misguided, and the machine perfected the mistake at scale</li>\n<li><strong>Wrong thing, wrong way</strong>: both failed, usually visibly enough to be caught first</li>\n</ul>\n<p>Every outcome traces back to a human. This was a consequence of computers, for better or worse, doing precisely what they’re instructed to do. It was also a consequence of humans, despite being ever more collectively corralled by sociocultural systems, creatively subverting instructions. The machine contributed nothing but fidelity, so the human was the only place judgement, and therefore error, could live.</p>\n<p>Nowadays, a whole new class of machines isn’t so deterministic, and societal mechanisms of control are fracturing and fraying in step. Probabilistic machines reopen the possibility space with two outcomes the old one ruled out:</p>\n<ul>\n<li><strong>Human wrong, machine right</strong>: the machine catches the human’s mistake and does what was meant rather than what was said</li>\n<li><strong>Human right, machine wrong</strong>: the machine drifts, reinterprets or confabulates, and the error starts with it rather than the human</li>\n</ul>\n<p>Once machines can subvert instructions too, outcomes no longer trace back to a single human. Judgement and error are distributed, and attribution, the thing credit and blame both depend on, starts to dissolve.</p>\n<p>It gets more complicated, though, because the action is migrating to scales humans can’t inhabit.</p>\n<p>At one end, high-frequency trading and agent swarms create miniaturised economies too fast and too small for humans to make sense of. At the other, exocapitalism is subliming capital from reality. Meanwhile, we’re having to work out Long Now problems we’ve been putting off for generations, problems too slow and too large for any one lifetime.</p>\n<p>Lived experience has been here before. When steam replaced sail, the sailor’s hard-won feel for wind and weather gave way to stokers shovelling coal below deck. The ship still crossed the ocean and life aboard went on, but the people aboard no longer had a feel for how.</p>\n<p>The upside is that all that complexity can be compressed into protocols. It just takes thought, human or otherwise. The cost is that human lived experience gets well and truly de-centred. It isn’t erased, though. It goes on, as rich as ever, but as one integral thread in a many-dimensional fabric of reality rather than the loom it’s woven on. For as long as we insist on being the centre, the fullest upside of protocols stays locked away.</p>\n<hr />\n<p>A few critical questions remain in order to truly describe the nature of protocol thinking for product people. A running list for reference:</p>\n<ul>\n<li>The enduring place and value of product thinking</li>\n<li>The most appropriate <a href=\"https://msweet.net/notes/112-pearl-falcons\">metaphors</a> for protocol thinking</li>\n<li>Protocol thinking vs projects, platforms and other “P” paradigms</li>\n<li>Formal verification and proof interconnection</li>\n<li>The historical context of protocols for product (e.g. ZIRP era exit, World Machines)</li>\n</ul>\n<p>Ongoing gratitude to the <a href=\"https://blygger.org/contributors/\">Blygger contributors</a> for inspiring the above.</p>\n<p>Also of note: soon to be a <a href=\"https://protocol-institute.org/projects/project/?slug=protocol-thinking-for-product-people\">book</a> as part of the Protocol Institute’s 2026 <a href=\"https://protocol-institute.org/events/pibowrimo-2026/\">PiBoWriMo</a>.</p>",
  "content_hash": "sha256:45214b9679938d32c6e18c441adb9a7b718d4b31934aade9dfc26294229cb19f"
}
