{ "$schema": "https://ndcodex.com/schemas/object/v1.json", "id": "codex://object/the-shape-of-the-work", "archive_id": "the-shape-of-the-work", "slug": "the-shape-of-the-work", "url": "https://ndcodex.com/codex/the-shape-of-the-work/", "type": "codex", "object_form": null, "title": "The Shape of the Work", "summary": "A roundup of Design Intelligence, File Zero, Mulcher, Core, Torn Frames, poetry, and research—and how learning moves between them.", "content_text": "Lately, I have been building the work and the conditions that let the work continue.\n\nWebsites, design studies, poems, torn paper, research, archives. Alongside them: the systems that help me find something again, carry a decision forward, collaborate with an agent, or recognize that a question from one project has wandered into another.\n\nSeveral things are developing at once.\n\nDesign Intelligence. File Zero. Mulcher. A shared operating layer I have been calling Core. Torn Frames. The writing here at ndcodex. An expanding inquiry into spiritual mechanics and the broader question of how things work.\n\nThey are not all the same project.\n\nBut their relationships are becoming easier to see.\n\nA question keeps appearing underneath them:\n\nHow do we make what we learn available to what we do next?\n\nNot just stored.\n\nAvailable.\n\nSomething we can find, understand, question, use, revise, and return to the work.\n\nThat is a good place to begin this roundup.\n\nDesign that remembers\n\nA design project produces more than its deliverables.\n\nThere is the finished website, certainly. But there is also everything discovered while making it: why one arrangement worked, where the navigation became confusing, what a particular type relationship accomplished, which assumptions survived contact with actual content.\n\nI want more of that learning to remain accessible.\n\nThat is the work taking shape around Design Intelligence.\n\nThe ambition is not simply a larger collection of references. I already know how to collect things. The harder question is what happens after collection.\n\nCan an observation become useful intelligence?\n\nCan a successful pattern retain the context that made it successful?\n\nCan a future project inherit a lesson without inheriting an entire previous project’s appearance?\n\nAn editorial study, for example, might begin with the relationship between an eyebrow, headline, introduction, paragraph, and action. But the useful learning is not limited to their sizes. It includes sequence, emphasis, pacing, content length, image behavior, and the circumstances under which the arrangement starts to fail.\n\nI want to preserve those relationships.\n\nNot just the screenshot.\n\nThe screenshot shows what happened. The study should help explain what was working.\n\nDesign Intelligence is my attempt to give that learning somewhere to accumulate without prematurely declaring every discovery a rule.\n\nFile Zero: a place to begin\n\nWithin that work, File Zero is becoming the proposed starting environment for design in Figma.\n\nNot a finished aesthetic waiting to be applied to everything.\n\nA prepared place to work.\n\nResponsive starting structures. Adaptable parts. Clear naming. Space for references and studies. Ways to distinguish an experiment from something ready for review. Conventions that help a human or an agent enter the file and understand what is happening.\n\nI do not want to solve the same basic setup problems every time a new project begins.\n\nI also do not want efficiency to become an excuse for making the same website repeatedly.\n\nThe distinction matters:\n\nReuse the support structure. Keep the design available to discovery.\n\nOne useful frame has been the file as both shop floor and exhibit.\n\nOn the shop floor, the work can be unfinished. Studies sit beside alternatives. Questions remain visible. Something can be promising without being approved.\n\nThe exhibit is where the work becomes legible to someone who has not been standing beside the machinery all day.\n\nWhat are they looking at?\n\nWhat decision is being requested?\n\nWhat is settled?\n\nWhat is still open?\n\nA review should not require someone to excavate the designer’s entire working memory.\n\nThe same applies to feedback. I want comments and decisions to become part of the project’s retained intelligence, rather than a conversation that disappears once the next screen is drawn.\n\nAnd underneath all of this is the question of when to harden.\n\nWhen does a useful arrangement deserve to become a reusable component?\n\nWhen is a preference mature enough to become a convention?\n\nWhen should a promising idea remain a study?\n\nPart of the operator’s job is knowing when to leave something soft.\n\nMulcher and the front door\n\nBefore something becomes intelligence, it often arrives as a fragment.\n\nA screenshot. A paragraph. A strange observation. A link with three words attached. A question that does not yet know which discipline it belongs to.\n\nMulcher is the developing intake and processing project around that material.\n\nOne of the simplest decisions has also been one of the most useful: give incoming material a clear front door.\n\nThe Jar is an inbox outside the curated vault.\n\nDropping something into it does not make that thing important, verified, or permanent. It means the material has somewhere to land without requiring me to finish thinking about it first.\n\nSelection and ingestion are separate acts.\n\nThat is where a receipt becomes meaningful: what was selected, where it went, what was extracted, and what remains unresolved.\n\nCollection is not endorsement.\n\nExtraction is not verification.\n\nAnd a compelling observation does not become established knowledge merely because an agent can write a beautiful paragraph around it.\n\nThe aim is to move from accumulated material toward usable understanding while keeping the path between them visible.\n\nSome fragments will become studies. Some will support an existing question. Some will expose a new one.\n\nSome should simply remain fragments.\n\nThe person beneath the systems\n\nThese projects also keep revealing the need for a shared operating layer.\n\nI have been calling that layer Core.\n\nThe basic distinction is that my domains of work are not the same thing as the infrastructure that supports them.\n\nFreelance work has its own responsibilities. Art has its own purposes. Personal projects have theirs. This journal is not merely a marketing department for any of them.\n\nShared infrastructure can serve those domains without owning them.\n\nThat includes the practical material: project orientation, working conventions, device roles, storage, handoffs, archives, and the information an agent needs to collaborate without requiring a complete reintroduction every time.\n\nAn agent needs more than a task.\n\nIt needs to know what kind of work it has entered, what it is allowed to change, what counts as evidence, what remains undecided, and where human judgment belongs.\n\nI am trying to make those conditions more explicit.\n\nNot to remove myself from the work.\n\nTo spend less of myself repeatedly explaining where the work is.\n\nSome of this is already being exercised in live projects. Some is still proposed architecture. I want that distinction to remain visible.\n\nA diagram of an operating system is not proof that it operates.\n\nEventually, the thing has to help on an ordinary day.\n\nTorn Frames: the object keeps its history\n\nThe art practice offers another version of these questions, but it deserves to remain art.\n\nTorn Frames is developing around torn paper, salvage, singular objects, and the archive that can hold them.\n\nHere, preservation is not simply a matter of keeping the file.\n\nWhat is the object?\n\nWhere did the material come from?\n\nWhat happened to it?\n\nWhich marks belong to its previous life, and which belong to the act of making this one?\n\nHow much do I know, and where should the record say that I do not know?\n\nI am interested in documentation that gives an object presence without pretending to explain it away.\n\nA title, a date, a material description, a condition note, a fragment of provenance. These can create a way into the work without closing off the encounter.\n\nThe frame matters too.\n\nIt can contain the image, interrupt it, or become part of what the image is doing. The boundary is not always neutral.\n\nThere is a relationship here to interface design and information architecture. But the relationship does not make the artwork a design exercise.\n\nThe object does not owe the system a lesson.\n\nSometimes the system’s entire job is to let the object remain strange.\n\nLanguage is also a working surface\n\nThe poetry continues alongside the architecture.\n\nSometimes before it.\n\nA line arrives before there is a category for it. A phrase catches something the more formal language has been circling. Rhythm finds an opening that explanation could not.\n\nI work through fragments, line by line, and eventually through the larger act of stitching.\n\nSpelling, punctuation, capitalization, spacing, breaks. Every character can change the pressure of the thing.\n\nThis is part of why I keep returning to raw markdown. I want the text close enough to touch. I want to see what has actually been placed there.\n\nBut the writing is not just another material waiting to be processed.\n\nA poem is not an inefficient research note.\n\nIts uncertainty may be essential. Its contradiction may be the point. Its force may depend on refusing the tidy explanation I could produce afterward.\n\nUsing language to unbury me.\n\nThat phrase names something the diagrams do not.\n\nThere is a person underneath the systems, and language is one of the ways that person remains present.\n\nNot everything I write needs to become doctrine, a framework, or a reusable asset.\n\nSome of it needs to speak.\n\nSome of it needs to make a sound large enough to stand inside.\n\nThe shape of things\n\nThe spiritual mechanics study belongs within this larger practice of inquiry.\n\nIt is one branch of a broader fascination:\n\nHow does a thing work?\n\nWhat are its structures, postures, states, functions, exchanges, thresholds, and failures?\n\nWhat does it do?\n\nWhat do people say it does?\n\nWhat can we observe, and what are we interpreting?\n\nIn the spiritual study, that means examining proposed spiritual anatomies, practices, experiences, symbols, and forms of authority. It means placing theological language beside other ways of studying human life without pretending they are interchangeable.\n\nBut the same habit of inquiry extends into design, behavior, culture, institutions, technology, ethics, and justice.\n\nA navigation system proposes a way through something.\n\nAn institution organizes participation and authority.\n\nA ritual gives an action a setting and a sequence.\n\nA poem arranges attention through language.\n\nThese are not equivalent mechanisms. They are different places where questions about structure and behavior become useful.\n\nI am not trying to force them into one master explanation.\n\nI am trying to improve my ability to notice what is happening.\n\nContinued understanding, rather than full knowledge.\n\nThe distinction keeps the work open.\n\nCross-informed, not collapsed\n\nOne of the most promising questions has also been one of the simplest:\n\nHow do these all inform one another?\n\nWhat might an editorial study reveal about the presentation of research?\n\nWhat might archive practices teach a design system about provenance?\n\nWhat might poetry expose about pacing, emphasis, and omission?\n\nWhat might an inquiry into authority help me notice about permissions in an agentic workflow?\n\nThese are invitations to investigate, not declarations that a correspondence has been proven.\n\nThe individual stores need to retain their sovereignty.\n\nA theological interpretation should not quietly become a scientific finding. A design preference should not acquire the authority of a universal law. A powerful metaphor should remain distinguishable from evidence.\n\nI want to mark the crossings without erasing the borders.\n\nThat suggests a useful kind of record: not only what we know, but which relationships are worth examining next.\n\nA store of golden queries.\n\nQuestions specific enough to pursue and open enough to reveal something unexpected.\n\nSometimes the next useful act is not a grand research expedition.\n\nIt is a small, well-aimed inquiry that reveals whether a thread is worth following.\n\nThe return is part of the work\n\nAll of this needs an ethic beyond accumulation.\n\nMore files, more agents, more categories, and more automation are not sufficient evidence that something useful is happening.\n\nThe system has to return something.\n\nA clearer decision. A stronger piece of work. Less repeated setup. An observation that would otherwise have been lost. A handoff another person can actually use.\n\nRoom to think.\n\nRoom to stop.\n\nI do not want infrastructure that recruits the entire practice into maintaining the infrastructure.\n\nThe machinery serves the work.\n\nThe work does not exist to keep the machinery busy.\n\nThe creative ethic I keep returning to is:\n\nDwell. Perceive. Discern. Make. Return.\n\nReturn means bringing something back from the encounter.\n\nIt means allowing the result to improve the conditions of the next attempt.\n\nIt also means returning to the people affected by the work, to the question that justified the effort, and to the world beyond the system.\n\nA loop that never returns anything has become a demand.\n\nWhat this place is for\n\nndcodex is where I can keep a public record of these developing relationships.\n\nNot only finished explanations.\n\nField notes. Poems. Studies. Objects. Questions. Working models. Revisions. Occasional evidence that an earlier conviction needed to be dismantled.\n\nI do not need everything to resolve into one identity before it can belong here.\n\nThe designer can remain a designer.\n\nThe poet can interrupt.\n\nThe artwork can refuse the framework.\n\nThe research can correct the story.\n\nThe person can decide that the system needs to be smaller.\n\nWhat connects these things is not that they all perform the same function.\n\nIt is the practice of paying attention, making something from the encounter, and remaining available to what the result teaches.\n\nSo that is the current shape of the work.\n\nA design practice learning how to remember.\n\nAn archive learning how to preserve context.\n\nA collaboration learning how to carry judgment without pretending to replace it.\n\nA collection of questions learning where to look next.\n\nAnd language, still moving through all of it.\n\nSome of the structure is becoming clear.\n\nSome of it is still a pencil mark.\n\nThe next piece of work gets to tell me what needs redrawing.", "content_markdown": "# The Shape of the Work\n\n![An open frame of torn paper rests among overlapping fragments, graphite grids, and charcoal marks on a warm paper surface.]()\n\nLately, I have been building the work and the conditions that let the work continue.\n\nWebsites, design studies, poems, torn paper, research, archives. Alongside them: the systems that help me find something again, carry a decision forward, collaborate with an agent, or recognize that a question from one project has wandered into another.\n\nSeveral things are developing at once.\n\nDesign Intelligence. File Zero. Mulcher. A shared operating layer I have been calling Core. Torn Frames. The writing here at ndcodex. An expanding inquiry into spiritual mechanics and the broader question of how things work.\n\nThey are not all the same project.\n\nBut their relationships are becoming easier to see.\n\nA question keeps appearing underneath them:\n\n**How do we make what we learn available to what we do next?**\n\nNot just stored.\n\nAvailable.\n\nSomething we can find, understand, question, use, revise, and return to the work.\n\nThat is a good place to begin this roundup.\n\n## Design that remembers\n\nA design project produces more than its deliverables.\n\nThere is the finished website, certainly. But there is also everything discovered while making it: why one arrangement worked, where the navigation became confusing, what a particular type relationship accomplished, which assumptions survived contact with actual content.\n\nI want more of that learning to remain accessible.\n\nThat is the work taking shape around **Design Intelligence**.\n\nThe ambition is not simply a larger collection of references. I already know how to collect things. The harder question is what happens after collection.\n\nCan an observation become useful intelligence?\n\nCan a successful pattern retain the context that made it successful?\n\nCan a future project inherit a lesson without inheriting an entire previous project’s appearance?\n\nAn editorial study, for example, might begin with the relationship between an eyebrow, headline, introduction, paragraph, and action. But the useful learning is not limited to their sizes. It includes sequence, emphasis, pacing, content length, image behavior, and the circumstances under which the arrangement starts to fail.\n\nI want to preserve those relationships.\n\nNot just the screenshot.\n\n**The screenshot shows what happened. The study should help explain what was working.**\n\nDesign Intelligence is my attempt to give that learning somewhere to accumulate without prematurely declaring every discovery a rule.\n\n## File Zero: a place to begin\n\nWithin that work, **File Zero** is becoming the proposed starting environment for design in Figma.\n\nNot a finished aesthetic waiting to be applied to everything.\n\nA prepared place to work.\n\nResponsive starting structures. Adaptable parts. Clear naming. Space for references and studies. Ways to distinguish an experiment from something ready for review. Conventions that help a human or an agent enter the file and understand what is happening.\n\nI do not want to solve the same basic setup problems every time a new project begins.\n\nI also do not want efficiency to become an excuse for making the same website repeatedly.\n\nThe distinction matters:\n\n**Reuse the support structure. Keep the design available to discovery.**\n\nOne useful frame has been the file as both **shop floor and exhibit**.\n\nOn the shop floor, the work can be unfinished. Studies sit beside alternatives. Questions remain visible. Something can be promising without being approved.\n\nThe exhibit is where the work becomes legible to someone who has not been standing beside the machinery all day.\n\nWhat are they looking at?\n\nWhat decision is being requested?\n\nWhat is settled?\n\nWhat is still open?\n\nA review should not require someone to excavate the designer’s entire working memory.\n\nThe same applies to feedback. I want comments and decisions to become part of the project’s retained intelligence, rather than a conversation that disappears once the next screen is drawn.\n\nAnd underneath all of this is the question of **when to harden**.\n\nWhen does a useful arrangement deserve to become a reusable component?\n\nWhen is a preference mature enough to become a convention?\n\nWhen should a promising idea remain a study?\n\nPart of the operator’s job is knowing when to leave something soft.\n\n## Mulcher and the front door\n\nBefore something becomes intelligence, it often arrives as a fragment.\n\nA screenshot. A paragraph. A strange observation. A link with three words attached. A question that does not yet know which discipline it belongs to.\n\n**Mulcher** is the developing intake and processing project around that material.\n\nOne of the simplest decisions has also been one of the most useful: give incoming material a clear front door.\n\nThe **Jar** is an inbox outside the curated vault.\n\nDropping something into it does not make that thing important, verified, or permanent. It means the material has somewhere to land without requiring me to finish thinking about it first.\n\nSelection and ingestion are separate acts.\n\nThat is where a receipt becomes meaningful: what was selected, where it went, what was extracted, and what remains unresolved.\n\nCollection is not endorsement.\n\nExtraction is not verification.\n\nAnd a compelling observation does not become established knowledge merely because an agent can write a beautiful paragraph around it.\n\nThe aim is to move from accumulated material toward usable understanding while keeping the path between them visible.\n\nSome fragments will become studies. Some will support an existing question. Some will expose a new one.\n\nSome should simply remain fragments.\n\n## The person beneath the systems\n\nThese projects also keep revealing the need for a shared operating layer.\n\nI have been calling that layer **Core**.\n\nThe basic distinction is that my domains of work are not the same thing as the infrastructure that supports them.\n\nFreelance work has its own responsibilities. Art has its own purposes. Personal projects have theirs. This journal is not merely a marketing department for any of them.\n\nShared infrastructure can serve those domains without owning them.\n\nThat includes the practical material: project orientation, working conventions, device roles, storage, handoffs, archives, and the information an agent needs to collaborate without requiring a complete reintroduction every time.\n\nAn agent needs more than a task.\n\nIt needs to know what kind of work it has entered, what it is allowed to change, what counts as evidence, what remains undecided, and where human judgment belongs.\n\nI am trying to make those conditions more explicit.\n\nNot to remove myself from the work.\n\nTo spend less of myself repeatedly explaining where the work is.\n\nSome of this is already being exercised in live projects. Some is still proposed architecture. I want that distinction to remain visible.\n\nA diagram of an operating system is not proof that it operates.\n\nEventually, the thing has to help on an ordinary day.\n\n## Torn Frames: the object keeps its history\n\nThe art practice offers another version of these questions, but it deserves to remain art.\n\n**Torn Frames** is developing around torn paper, salvage, singular objects, and the archive that can hold them.\n\nHere, preservation is not simply a matter of keeping the file.\n\nWhat is the object?\n\nWhere did the material come from?\n\nWhat happened to it?\n\nWhich marks belong to its previous life, and which belong to the act of making this one?\n\nHow much do I know, and where should the record say that I do not know?\n\nI am interested in documentation that gives an object presence without pretending to explain it away.\n\nA title, a date, a material description, a condition note, a fragment of provenance. These can create a way into the work without closing off the encounter.\n\nThe frame matters too.\n\nIt can contain the image, interrupt it, or become part of what the image is doing. The boundary is not always neutral.\n\nThere is a relationship here to interface design and information architecture. But the relationship does not make the artwork a design exercise.\n\nThe object does not owe the system a lesson.\n\nSometimes the system’s entire job is to let the object remain strange.\n\n## Language is also a working surface\n\nThe poetry continues alongside the architecture.\n\nSometimes before it.\n\nA line arrives before there is a category for it. A phrase catches something the more formal language has been circling. Rhythm finds an opening that explanation could not.\n\nI work through fragments, line by line, and eventually through the larger act of stitching.\n\nSpelling, punctuation, capitalization, spacing, breaks. Every character can change the pressure of the thing.\n\nThis is part of why I keep returning to raw markdown. I want the text close enough to touch. I want to see what has actually been placed there.\n\nBut the writing is not just another material waiting to be processed.\n\nA poem is not an inefficient research note.\n\nIts uncertainty may be essential. Its contradiction may be the point. Its force may depend on refusing the tidy explanation I could produce afterward.\n\n**Using language to unbury me.**\n\nThat phrase names something the diagrams do not.\n\nThere is a person underneath the systems, and language is one of the ways that person remains present.\n\nNot everything I write needs to become doctrine, a framework, or a reusable asset.\n\nSome of it needs to speak.\n\nSome of it needs to make a sound large enough to stand inside.\n\n## The shape of things\n\nThe spiritual mechanics study belongs within this larger practice of inquiry.\n\nIt is one branch of a broader fascination:\n\n**How does a thing work?**\n\nWhat are its structures, postures, states, functions, exchanges, thresholds, and failures?\n\nWhat does it do?\n\nWhat do people say it does?\n\nWhat can we observe, and what are we interpreting?\n\nIn the spiritual study, that means examining proposed spiritual anatomies, practices, experiences, symbols, and forms of authority. It means placing theological language beside other ways of studying human life without pretending they are interchangeable.\n\nBut the same habit of inquiry extends into design, behavior, culture, institutions, technology, ethics, and justice.\n\nA navigation system proposes a way through something.\n\nAn institution organizes participation and authority.\n\nA ritual gives an action a setting and a sequence.\n\nA poem arranges attention through language.\n\nThese are not equivalent mechanisms. They are different places where questions about structure and behavior become useful.\n\nI am not trying to force them into one master explanation.\n\nI am trying to improve my ability to notice what is happening.\n\n**Continued understanding, rather than full knowledge.**\n\nThe distinction keeps the work open.\n\n## Cross-informed, not collapsed\n\nOne of the most promising questions has also been one of the simplest:\n\n**How do these all inform one another?**\n\nWhat might an editorial study reveal about the presentation of research?\n\nWhat might archive practices teach a design system about provenance?\n\nWhat might poetry expose about pacing, emphasis, and omission?\n\nWhat might an inquiry into authority help me notice about permissions in an agentic workflow?\n\nThese are invitations to investigate, not declarations that a correspondence has been proven.\n\nThe individual stores need to retain their sovereignty.\n\nA theological interpretation should not quietly become a scientific finding. A design preference should not acquire the authority of a universal law. A powerful metaphor should remain distinguishable from evidence.\n\nI want to mark the crossings without erasing the borders.\n\nThat suggests a useful kind of record: not only what we know, but which relationships are worth examining next.\n\nA store of **golden queries**.\n\nQuestions specific enough to pursue and open enough to reveal something unexpected.\n\nSometimes the next useful act is not a grand research expedition.\n\nIt is a small, well-aimed inquiry that reveals whether a thread is worth following.\n\n## The return is part of the work\n\nAll of this needs an ethic beyond accumulation.\n\nMore files, more agents, more categories, and more automation are not sufficient evidence that something useful is happening.\n\nThe system has to return something.\n\nA clearer decision. A stronger piece of work. Less repeated setup. An observation that would otherwise have been lost. A handoff another person can actually use.\n\nRoom to think.\n\nRoom to stop.\n\nI do not want infrastructure that recruits the entire practice into maintaining the infrastructure.\n\nThe machinery serves the work.\n\nThe work does not exist to keep the machinery busy.\n\nThe creative ethic I keep returning to is:\n\n**Dwell. Perceive. Discern. Make. Return.**\n\nReturn means bringing something back from the encounter.\n\nIt means allowing the result to improve the conditions of the next attempt.\n\nIt also means returning to the people affected by the work, to the question that justified the effort, and to the world beyond the system.\n\nA loop that never returns anything has become a demand.\n\n## What this place is for\n\nndcodex is where I can keep a public record of these developing relationships.\n\nNot only finished explanations.\n\nField notes. Poems. Studies. Objects. Questions. Working models. Revisions. Occasional evidence that an earlier conviction needed to be dismantled.\n\nI do not need everything to resolve into one identity before it can belong here.\n\nThe designer can remain a designer.\n\nThe poet can interrupt.\n\nThe artwork can refuse the framework.\n\nThe research can correct the story.\n\nThe person can decide that the system needs to be smaller.\n\nWhat connects these things is not that they all perform the same function.\n\nIt is the practice of paying attention, making something from the encounter, and remaining available to what the result teaches.\n\nSo that is the current shape of the work.\n\nA design practice learning how to remember.\n\nAn archive learning how to preserve context.\n\nA collaboration learning how to carry judgment without pretending to replace it.\n\nA collection of questions learning where to look next.\n\nAnd language, still moving through all of it.\n\nSome of the structure is becoming clear.\n\nSome of it is still a pencil mark.\n\nThe next piece of work gets to tell me what needs redrawing.", "author": { "name": "Nathan Davis", "designation": "Archive Operator", "role": "Archive Operator", "avatar": "/media/people/nathan-davis.jpg" }, "contributors": [ { "name": "Nathan Davis", "designation": "Archive Operator", "role": "Archive Operator", "avatar": "/media/people/nathan-davis.jpg" } ], "date_published": "2026-09-05T13:31:09.539Z", "date_modified": "2026-09-05T13:31:09.539Z", "status": "published", "visibility": "public", "language": "en-US", "axes": { "scale": "macro", "depth": "recursive", "focus": "witness", "function": "revelatory" }, "themes": [ "creative practice", "retained learning", "provenance", "cross-disciplinary inquiry", "creative infrastructure" ], "constellations": [], "tags": [ "creative practice", "design intelligence", "archives", "poetry", "research", "human-agent collaboration", "retained learning", "provenance", "cross-disciplinary inquiry", "creative infrastructure" ], "keywords": [ "Codex", "creative practice", "retained learning", "provenance", "cross-disciplinary inquiry", "creative infrastructure", "design intelligence", "archives", "poetry", "research", "human-agent collaboration" ], "relations": [], "media": [ { "kind": "image", "src": "/media/codex/the-shape-of-the-work/the-shape-of-the-work-hero.webp", "role": "hero", "alt": "An open frame of torn paper rests among overlapping fragments, graphite grids, and charcoal marks on a warm paper surface.", "caption": "AI-generated editorial illustration for The Shape of the Work." } ], "capture": null }