<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Itsuki Mokuda on Medium]]></title>
        <description><![CDATA[Stories by Itsuki Mokuda on Medium]]></description>
        <link>https://medium.com/@itsuki.mokuda?source=rss-05bba93ad562------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*Y9vr_o3wbW8rFc3S0j_7vw.png</url>
            <title>Stories by Itsuki Mokuda on Medium</title>
            <link>https://medium.com/@itsuki.mokuda?source=rss-05bba93ad562------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 23:07:26 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@itsuki.mokuda/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[An Invitation]]></title>
            <link>https://medium.com/@itsuki.mokuda/an-invitation-21aa0f55a0a3?source=rss-05bba93ad562------2</link>
            <guid isPermaLink="false">https://medium.com/p/21aa0f55a0a3</guid>
            <category><![CDATA[art]]></category>
            <category><![CDATA[japan]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[contemporary-art]]></category>
            <category><![CDATA[philosophy]]></category>
            <dc:creator><![CDATA[Itsuki Mokuda]]></dc:creator>
            <pubDate>Sun, 08 Nov 2026 09:31:30 GMT</pubDate>
            <atom:updated>2026-11-08T09:50:23.912Z</atom:updated>
            <content:encoded><![CDATA[<h3>A Question Worth Seeing</h3><p><em>Drifting Contours</em> is a project built around a single question: when a self passes through a system — trained, generated, selected, copied, and fed back in — what, if anything, remains?</p><p>This December, some of that process becomes visible in person.</p><h3>What You Will See</h3><p><a href="https://itsuki-mokuda.netlify.app/info/info_practice.mp4">https://itsuki-mokuda.netlify.app/info/info_practice.mp4</a></p><p>The works on display are hand-painted — made by Mokuda tracing a single image chosen from among multiple AI-generated outputs, each generated from Mokuda’s own past work. That act of choosing — selecting the one that feels most like “me” — is where the artist’s own assumptions and self-image press most visibly into the process, before the hand has even touched the canvas. Each piece is the material result of that cycle: not a collaboration in the usual sense, but a question made physical.</p><p>Showing alongside is work by Sakamoto, — layered paintings on transparent acrylic, built up through depth and light.</p><h3>Exhibition Details</h3><p><strong>Two-Person Exhibition</strong> Akiko Sakamoto × Itsuki Mokuda</p><p><a href="https://gallerykanon.boo.jp/access.html">Gallery Kanon</a>, Nihonbashi, Tokyo Dec 6 — Dec 12, 2026 12:30–18:30 (Last day until 17:00) Free admission</p><p>If you are in Tokyo (Japan) that week and the question interests you, we would be glad to see you there.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*CCfibwC07z4391Yz0mdkEA.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=21aa0f55a0a3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Phenomenological Work 001]]></title>
            <link>https://medium.com/@itsuki.mokuda/phenomenological-work-001-36402045397a?source=rss-05bba93ad562------2</link>
            <guid isPermaLink="false">https://medium.com/p/36402045397a</guid>
            <category><![CDATA[art]]></category>
            <category><![CDATA[philosophy]]></category>
            <category><![CDATA[creativity]]></category>
            <category><![CDATA[individuality]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Itsuki Mokuda]]></dc:creator>
            <pubDate>Tue, 14 Jul 2026 10:27:12 GMT</pubDate>
            <atom:updated>2026-07-14T10:57:30.568Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JoghQVxKaru1Q2ArFOUDLw.png" /></figure><h4>One image, born from feeding my past paintings into an AI. Is this me — or something that isn’t?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RPMytzV_vgXQ9zNoBhnTew.png" /><figcaption>[More on the premise here →<a href="https://medium.com/@itsuki.mokuda/about">https://medium.com/@itsuki.mokuda/about</a>]</figcaption></figure><p>For a long time, I believed there was something at the centre of the self — a core that could be expressed through making. I pursued it. The deeper the pursuit, the more that centre and any clear form began to dissolve and disappear, until I arrived at a single question: <em>What if there is nothing there to express?</em></p><p>From that point, I spent over 15 years working across traditional Japanese painting (Nihonga), botanical and figurative works in acrylic, abstraction, and metal casting using low-melting alloys — not in pursuit of a unified style, but to ensure that technical limitation would never stand in the way once that question finally found its form.</p><p><em>Drifting Contours</em> is that form. The first body of work in which the question itself has become the subject.</p><p>Given that premise, the first time I generated an image with AI, my honest reaction wasn’t resistance. It was something closer to surrender. Part of it was luck — the prompt happened to land well — but what I felt was: <em>what a beautiful image. Is this simply better than what I could make? And if so, what does it mean to keep painting at all?</em> Not disappointment in myself, exactly. More like recognizing that something else might be stronger than me at giving form to what I wanted to see.</p><p>The next thought followed quickly: if using this in my own work would raise its quality, why not? But that led straight into a different discomfort — wasn’t that closer to cheating on a test than to skill? I saw the same hesitation, or outright rejection, from people around me. So I didn’t use it. Not in the work itself.</p><p>But sitting with that hesitation over and over, I remembered something about my own position: I had never actually believed in a “pure individuality,” untouched by artifice, that belonged to the artist alone. I’d been skeptical of that idea for a long time — long before AI entered the picture.</p><p>Which meant the question became: if I don’t believe that fixed, protectable core exists, then who is better positioned to test it out in the open than I am? I’ve simply made enough over the years to feed into the process. I have no real attachment to “having a style” to defend. And if anything, I want to hold the question up and ask it plainly, rather than keep it private. Those three conditions lined up almost too neatly. I concluded I was, if anyone was, the right subject for this experiment.</p><h3>On the Title: Drifting Contours</h3><p>A contour marks the boundary between one thing and another. In this project, however, the boundary between “what began with me” and “what passed through AI” does not settle anywhere — it drifts, shifting and transforming as it goes. A contour that cannot be fixed: this reflects the nature of the project as a whole, and points to the uncertainty of “the self” itself.</p><h3>On the Title: Phenomenological Work</h3><p>Phenomenological refers to a disposition of attending to what presents itself — not reducing it to background, interpretation, or intention, but meeting it as it is. Whether or not you know the process, the request is the same: let what is in front of you be seen on its own terms. The numbering indicates that this experiment is still ongoing.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=36402045397a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Recognizable and Unfamiliar at the Same Time]]></title>
            <link>https://medium.com/@itsuki.mokuda/recognizable-and-unfamiliar-at-the-same-time-4a187d60fdb4?source=rss-05bba93ad562------2</link>
            <guid isPermaLink="false">https://medium.com/p/4a187d60fdb4</guid>
            <category><![CDATA[philosophy]]></category>
            <category><![CDATA[art]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[creativity]]></category>
            <category><![CDATA[painting]]></category>
            <dc:creator><![CDATA[Itsuki Mokuda]]></dc:creator>
            <pubDate>Mon, 13 Jul 2026 10:51:01 GMT</pubDate>
            <atom:updated>2026-07-13T10:51:01.100Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*W3vHZM4HtzX24Pq-gxOESA.png" /></figure><h4>One image, born from feeding my past paintings into an AI. Is this me — or something that isn’t?</h4><h3>Cycle One</h3><p>The first generated image came back, and I wrote this down before I could talk myself out of it:</p><p><em>What appears here resembles “my work” — and yet something in it resists being simply accepted as “my work.” An unidentifiable something, giving off an essential unease.</em></p><p>I wrote that sentence carefully, and left it deliberately ambiguous — “resists being accepted” could mean rejection, or it could mean something closer to a neutral, almost clinical observation. The clearest way I’ve found to describe it: the wariness you’d feel watching an impressionist do an impression of you. The mannerisms land, something is unmistakably being tracked — and yet if you put the two faces side by side, you’d say, objectively, they don’t really look alike. Recognizable and unfamiliar at the same time, without either canceling the other out.</p><p><em>Drifting Contours</em> is an ongoing project built on a simple, repeating loop: train an AI on my own paintings, let it generate new images, transcribe those images back into paint by hand, and feed the results into the next round of training. [More on the premise here →<a href="https://medium.com/@itsuki.mokuda/about">https://medium.com/@itsuki.mokuda/about</a>]</p><p>This is the record of the first time through that loop.</p><h3>What came back</h3><p>I trained the model on a body of work that doesn’t stick to one genre or style — I’ve never worked that way. So when the first outputs appeared, they carried recognizable fragments: a gesture, a palette, a way of handling a surface. But nothing added up to a single, nameable “my style.” Someone who knows my work might have looked at these images and not recognized them at all.</p><p>I’m still not sure whether that says something about the AI’s distillation, or about the fact that I never had a fixed style to distill in the first place.</p><h3>Transcribing without deciding</h3><p>Repainting the generated images by hand turned out to be closer to <em>shakyō</em> — a Buddhist practice, common in Zen training, of copying a sutra by hand, stroke by stroke, without adding personal interpretation — than to anything I’d call “painting from reference.” The point wasn’t to improve or interpret what the AI produced. It was to withhold judgment and simply transcribe, on the theory that whatever is essential might surface more clearly when I stop trying to author it.</p><p>I made exactly one category of exception: where the image broke basic physical logic — an extra finger, an anatomically impossible joint — and that breakage didn’t read as interesting or beautiful on its own terms, I corrected it. Everything else, however unfamiliar, I let stand.</p><h3>Curating as I go</h3><p>The process isn’t as linear as “generate ten, transcribe ten.” I curate at the generation stage too — if a batch of ten outputs doesn’t give me anything worth transcribing, I’ll generate several dozen more and pull one image out of that larger set. And the adjustments to what feeds back into the next round happen <em>during</em> that back-and-forth, not as a separate step done beforehand — some cuts and additions are deliberate, meant to keep the outputs from converging too quickly into a single look; others are instinctive, images I removed simply because they felt wrong, in a way I trusted without fully being able to explain.</p><p>Whether that instinct is meaningful data or just noise is, I think, part of what this project is testing.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4a187d60fdb4" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>