日本記者クラブの2026年6月8日の記者会見「人口減少時代を生きる」（12）で、日本総合研究所の主席研究員、藤波匠さんの講演を見た。非常に質の高い内容で、記録しておきたいと思った。
At the Japan National Press Club press conference on June 8, 2026, “Living in an Age of Population Decline” (No. “12), I watched a talk by Takumi Fujinami, principal researcher at the Japan Research Institute. The content was of very high quality, and I wanted to keep a record of it.

人口減少には月並みに関心があった。面白おかしく伝えるメディアの番組ではなく、現実的で信頼のおける情報に触れれば、次のようなことを考える手助けになると思った。
Like many people, I was interested in population decline. I thought that realistic, reliable information, rather than a media program that turns it into entertainment, would help me think about things like these.

• 人手不足や高齢化が言われる業界のうち、本当に問題が深刻なのはどの業界なのか。それを調べるときに、相関関係を考えられる。
  Among the industries said to face labor shortages and aging, which ones are really in serious trouble? When I look into that, I can think about the correlations.
• 日本の経済を引っ張っている会社や団体で、意思決定層が退くのは何年ごろになるのか。見通しをつけてみたい。
  At the companies and organizations driving Japan’s economy, around what year will the decision-makers step down? I want to get a sense of that.

会見は約1時間33分で、講演が約63分、残りが質疑だった。動画はYouTubeにあり、会見リポートは日本記者クラブのサイト（日本経済新聞の編集委員兼論説委員が執筆）にある。
The press conference ran about 1 hour 33 minutes: the talk was about 63 minutes and the rest was Q&A. The video is on YouTube, and the press conference report, written by a Nikkei staff editor who is also an editorial writer, is on the Japan National Press Club site.

私は専門家ではない。動画と自動生成の字幕を見て、スライドの数値を確かめながらまとめた。講演の全文ではなく、要点とデータの記録だ。スライドは載せず、数値を読み取って図を作り直した（グラフから目で読んだ値は、図にそう書いた）。元のスライドは動画の中でしか見られないので、本文の「スライドp.〇・動画〇:〇〇」には、スライドのページ番号と、そのスライドが映り始める動画の時刻を書いた。間違いがあればご指摘お願いします。
I’m not an expert. I watched the video and its auto-generated captions and wrote this while checking the numbers against the slides. It is a record of the main points and the data, not a full transcript. I did not reproduce the slides; I read the numbers and redrew the charts (values read by eye from a chart are marked as such in the figure). The original slides can only be seen in the video, so where the text says “slide N at m:ss in the video”, it gives the slide’s page number and the time in the video when that slide first appears. I would appreciate it if you could point out any mistakes.

目次 / Contents

• What Regional Revitalization Set Out to Do
• The Inflow to the Tokyo Area Was Far from Zero
• “There Are Jobs” Is Not the Same as “There Are Good Jobs”
• Migrants Rose, but the Net Outflow Did Not Change
• The People Who Moved Were Middle-Aged and Older Men
• Women, Too, Move for Jobs
• New Jobs Gathered in the Tokyo Area and Kinki
• Young Women Went into Information and Communications
• Male Workers Keep Declining; the Growth Is in Women
• Changes Inside the Big Cities
• Fujinami’s Recommendations
• From the Q&A: Regional Companies Becoming Healthy
• My Thoughts
• Sources

要点（TL;DR。2026年6月8日の講演時点）
The short version (as of the June 8, 2026 talk)

• 地方創生は「雇用をしっかり作って、若い人が定着する」ことを目標に掲げていた。ただ、結果として目立ったのは移住促進だった。成果が数字で見えやすい一方で、質の高い雇用は作れなかった、と藤波さんは評価した。
  Regional revitalization set the goal of creating solid jobs so that young people would settle. But what stood out in the end was the promotion of migration. Its results are easy to see in numbers, while quality jobs were not created, Fujinami assessed.
• 山形県では、移住者が約44人から約335人に増えたのに、転出超過は約3,100〜4,550人の幅のまま変わらなかった。藤波さんは、転出超過の増減は移住者の増減とほぼ関係がなく、東京の景気に左右される、と整理した。
  In Yamagata Prefecture, the number of migrants rose from about 44 to about 335, yet the net outflow stayed within a range of about 3,100 to 4,550. Fujinami summarized that the rise and fall of the net outflow has almost no relation to the rise and fall of migrants and depends on Tokyo’s economy.
• 藤波さんは、人は雇用で動き、女性も同じだと見ている。2019〜24年の雇用の純増に対する寄与率（割合の比較で、増えた人数そのものではない）で見ると、東京圏と近畿で86.6%だった。全国の25〜34歳の女性では、増えた雇用への情報通信業の寄与率が約86%だった（小さな母数からの概算）。
  Fujinami sees people as moving for jobs, and women as no different. As a contribution to the net increase in jobs in 2019–24 (a comparison of shares, not of the number of jobs added), the Tokyo area and Kinki together came to 86.6%. For women aged 25–34 nationwide, the information and communications industry’s contribution to the growth in jobs was about 86% (an estimate from a small base).
• コロナ禍に東京圏から地方へ出ていった人は、男性が女性より多く、年齢別では45歳以上が増えた。若い女性の流入先には、ジェンダーギャップの小さい情報通信業の存在がある、と藤波さんは見ている。
  During the pandemic, more men than women left the Tokyo area for the regions, and by age the increase was in those aged 45 and over. Fujinami sees the information and communications industry, with its small gender gap, as a draw for young women moving in.
• 提言は、目の前の転出超過に焦らず、地方の企業が生産性と賃金を上げて高度人材を採ること。移住促進に金を投じても、リターンは小さいという。地方創生とは、結局、地方の企業が元気になることだ、と藤波さんは話した。
  The recommendation: don’t rush over the current net outflow, and let regional companies raise productivity and wages to hire skilled people. Putting money into migration promotion gives small returns, he said. Regional revitalization, Fujinami said, is in the end about regional companies becoming healthy.

What Regional Revitalization Set Out to Do

地方創生の出発点は、2015年の「まち・ひと・しごと創生総合戦略」だ。基本的な視点は3つで、東京一極集中の是正、若い世代の就労・結婚・子育ての希望の実現、地域の特性に即した地域課題の解決。目標は、地方における安定した雇用の創出、地方への新しいひとの流れ、若い世代の結婚・出産・子育ての希望の実現だった。
The starting point of regional revitalization was the 2015 “Town, People and Job Creation Comprehensive Strategy.” It had three basic viewpoints: correcting the concentration in Tokyo, realizing young people’s hopes for work, marriage and child-rearing, and solving local issues in line with each area’s character. Its goals were to create stable jobs in the regions, to create a new flow of people to the regions, and to fulfill young people’s hopes for marriage, childbirth and child-rearing.

藤波さんは当時、この計画を見て、東京一極集中の是正は難しいと思ったという。東京圏の転入超過は当時約12万人で、それを5年でゼロにする目標は、ハードルが高い。一方で、雇用をしっかり作って若い人が定着できる環境をつくる点と、地域が自ら課題を考えて計画を作り、国が競わせる点は、非常に良いと思ったと話した。
Looking at the plan at the time, Fujinami thought correcting the concentration in Tokyo would be difficult. The Tokyo area’s net inflow was then about 120,000 a year, and a goal of bringing it to zero in five years was a high hurdle. On the other hand, he said he found two points very good: building an environment where solid jobs let young people settle, and having each region think through its own issues and write its own plan, with the national government making them compete.

三大都市圏の長期の推移（住民基本台帳人口移動報告）を見ると、人が流れ込んでいるのは東京圏だけだ。東京圏の転入超過は、1960年代前半に約38万人あった。1975年前後に約5万人まで急に減り、そのあとは増減を繰り返しながら、2023〜25年は約11〜12万人で推移している。大阪圏と名古屋圏は、70年代半ば以降ほぼゼロだ（スライドp.4・動画4:38、グラフからの目測）。だから東京一極集中と見られがちで、その是正が政策の柱になった、と藤波さんは整理した。
Looking at the long-term trend of the three major metropolitan areas (Basic Resident Register migration report), only the Tokyo area is receiving people. Its net inflow was about 380,000 in the early 1960s. It fell sharply to about 50,000 around 1975, and since then has moved up and down, staying around 110,000 to 120,000 from 2023 to 2025. The Osaka and Nagoya areas have been close to zero since the mid-1970s (slide 4 at 4:38 in the video, values read by eye). That is why it tends to be seen as concentration in Tokyo, and correcting it became a pillar of the policy, Fujinami summarized.

The Inflow to the Tokyo Area Was Far from Zero

藤波さんは、東京圏の転入超過は、2019年までは地方創生の目標とは逆に増えた、と説明した。東京圏の転入超過（日本人）は、2015年に約12万人、2019年には約14.6万人まで増えた。コロナ禍の2021年に約8.1万人まで減ったが、2023年以降は約11.5〜12万人に戻り、2025年は約11.3万人だった（スライドp.7・動画11:40、目測）。コロナ禍で東京圏の吸引力が下がっても、講演の時点では東京回帰が起きている、というのが藤波さんの見立てだ。
Fujinami explained that, up to 2019, the Tokyo area’s net inflow grew in the opposite direction from the goal of regional revitalization. The Tokyo area’s net inflow (Japanese nationals) rose from about 120,000 in 2015 to about 146,000 in 2019. It fell to about 81,000 in the pandemic year 2021, but returned to about 115,000 to 120,000 from 2023, and was about 113,000 in 2025 (slide 7 at 11:40 in the video, values read by eye). Fujinami’s view is that even though the pandemic lowered the Tokyo area’s pull, a return to Tokyo was under way at the time of the talk.

藤波さんによると、2020年の緊急事態宣言が出たのは4月なので、それまでの1〜3月は、前年とほぼ同じ水準だった。コロナ禍がなければ、2019年と同じ水準か、さらに高かった可能性もあるという。藤波さんは、東京への人の流れを食い止めることはできなかった、と見ている。
According to Fujinami, the state of emergency in 2020 was declared in April, so January to March of that year was at about the same level as the year before. Without the pandemic, it might have stayed at the 2019 level or gone even higher. He believes the flow of people to Tokyo could not be stopped.

“There Are Jobs” Is Not the Same as “There Are Good Jobs”

地方創生の戦略は、雇用を作る重要性を指摘していた。ところが、現場では「仕事はあるのだから、人を呼び込めばいい」という発想になり、移住促進策に偏った、と藤波さんは整理した（スライドp.6・動画9:28）。実際、地域別の有効求人倍率は、2015年から2025年まで、地方圏が東京圏より高い。
The regional revitalization strategy pointed out the importance of creating jobs. But on the ground, the thinking became “there are jobs, so just bring people in,” and policy leaned toward promoting migration, Fujinami summarized (slide 6 at 9:28 in the video). In fact, the job-openings-to-applicants ratio by region was higher in regional Japan than in the Tokyo area throughout 2015 to 2025.

藤波さんが2015年ごろに地方へ講演に行くと、雇用が大事だという話には共感されたが、地方には仕事はある、むしろ人手不足だ、と言われたという。数字の上でも、地方圏のほうが高かった。ただ、藤波さんが言っていた雇用は、仕事があればいい、ということではなかった。より質の高い雇用、より高い賃金、大学を出た女性でも就職できる職場のことで、それが地方には足りなかった、と話した。
When Fujinami went to give talks in the regions around 2015, people agreed that jobs mattered, but told him that there were jobs in the regions, and that the real problem was a labor shortage. In the numbers too, regional Japan was higher. But the employment Fujinami had in mind was not simply having work. It meant higher-quality jobs, higher wages, and workplaces where even a woman with a university degree could find a job, and those were lacking in the regions, he said.

藤波さんは、行政が移住促進だけをやっていたわけではない、と断ったうえで、すぐに数字になるのが移住促進で、若い人を東京圏から引っ張る政策は目立ち、成果が上がって見える、と話した。質疑では、質問者が、移住説明会の来場者数のように、数字で実績が出しやすかったからではないか、という見方を示した。藤波さんは、講演の時点で10年を振り返り、残念な結果だったと評価し、質の高い雇用を作れなかったことが反省点だと述べた。
Fujinami made clear that local governments did not do only that, but said that migration promotion is what turns into numbers quickly, and that policies that pull young people from the Tokyo area stand out and make it look as if results are coming in. In the Q&A, a questioner suggested it may be because results are easy to show in numbers, such as attendance at migration briefings. Looking back on the ten years as of the talk, Fujinami judged the outcome disappointing and said the failure to create quality jobs is the point to reflect on.

Migrants Rose, but the Net Outflow Did Not Change

藤波さんは、山形県の例で、移住促進の限界を示した。山形県が特別なわけではなく、藤波さんがたまたま講演して、データを作った地域だという。山形県の移住者数（県や市町の制度に乗って入った人数）は、2016年度の約44人から、2024年度の約335人へ増えた。コロナ禍の2020〜21年度に特に伸びた。一方、転出超過は、2016年が約3,700人、2019年が約4,550人、2021年が約3,100人、2024年が約4,200人と、約3,100〜4,550人の幅で動いたままで、減っていない（スライドp.8・動画13:08、目測。移住者数は年度、転出超過数は暦年の値）。
Fujinami used Yamagata Prefecture to show the limits of promoting migration. Yamagata is not special; it is just a region where Fujinami happened to give a talk and so made the data. The number of migrants in Yamagata (people who came in through prefectural and municipal programs) rose from about 44 in fiscal 2016 to about 335 in fiscal 2024, with a particularly big rise in fiscal 2020–21. Yet the net outflow moved within a range of about 3,100 to 4,550 and did not shrink: about 3,700 in 2016, about 4,550 in 2019, about 3,100 in 2021, and about 4,200 in 2024 (slide 8 at 13:08 in the video, values read by eye; migrants are by fiscal year and the net outflow by calendar year).

藤波さんは、転出超過の増減は移住者の増減とほぼ関係がなく、東京の景気に大きく左右される、と説明した。東京で人手不足になると、地方から人が集まる。一方、山形に入ってくる人全体（制度を使った移住者以外も含む）の数は、日本の人口が減っていることもあり、ほぼ一貫して減っている。なお、移住者が最も伸びた2020〜21年度に、転出超過は約3,100人までいったん減ったが、2022〜24年は約3,800〜4,200人に戻った（図の点、目測）。
Fujinami explained that the rise and fall of the net outflow has almost no relation to the rise and fall of migrants, and depends heavily on Tokyo’s economy. When Tokyo faces a labor shortage, people are drawn from the regions. Meanwhile, the number of all people coming into Yamagata (including those other than migrants using the programs) has fallen almost consistently, partly because Japan’s population itself is shrinking. Note that in fiscal 2020–21, when migrants rose the most, the net outflow dipped to about 3,100, but in 2022–24 it returned to about 3,800 to 4,200 (points on the chart, read by eye).

石川県でも、藤波さんは移住者が転出超過をあまり動かせていないと説明した。移住者数は2016年の約890人から、2022年の約1,650人をピークに、2023〜24年は約1,310人になった。転出超過は、2016年の約800人から、2024年の約3,700人へ広がった（スライドp.9・動画15:46、目測）。2024年は能登半島地震の年で、この数字にはその影響もありうる（講演では触れていない、筆者の注）。
For Ishikawa Prefecture as well, he described migrants as not having moved the net outflow much. The number of migrants rose from about 890 in 2016 to a peak of about 1,650 in 2022, and was about 1,310 in 2023–24. The net outflow widened from about 800 in 2016 to about 3,700 in 2024 (slide 9 at 15:46 in the video, values read by eye). 2024 was the year of the Noto Peninsula earthquake, and this number may reflect its effect (not mentioned in the talk; my own note).

藤波さんは、移住者の獲得競争に意味がないとは言わなかった。お年寄りしか住んでいない限界集落に若い人が1人入ることは、地域にとって大きな成果になる。ただ、規模の大きい都市で移住者を取ろうとしても、出ていく人が多すぎて、影響は小さい。自治体の規模によっては、積極的に取る必要があったのかを、反省する必要があるという。
Fujinami did not say that competing to attract migrants is meaningless. When one young person moves into a marginal village where only elderly people live, it is a big achievement for the community. But in a large city, even trying to attract migrants has little effect, because far too many people are leaving. Depending on the size of the municipality, he said, we need to reflect on whether such an active policy was necessary.

The People Who Moved Were Middle-Aged and Older Men

コロナ禍では、地方移住、リモートワーク、二地域居住、デジタル田園都市国家構想などが盛んに語られた。では、実際に東京圏から地方へ出ていったのは誰か。年齢別に見ると、東京圏の転入超過の大半は15〜29歳で、30歳未満はコロナ禍の影響が軽微で、2023年には2019年の水準を超えた。変わったのは45歳以上で、もともと小さかった転出超過が、2020年から大きくなった（スライドp.10・動画16:58）。
During the pandemic, regional migration, remote work, dual-location living and the Digital Garden City Nation initiative were widely discussed. So who actually left the Tokyo area for the regions? By age, most of the Tokyo area’s net inflow is people aged 15 to 29, and those under 30 were only slightly affected by the pandemic, exceeding their 2019 level in 2023. What changed was those aged 45 and over, whose small net outflow grew larger from 2020 (slide 10 at 16:58 in the video).

男女別に見ると、差が出る。男性は、転出超過（ほとんどが45歳以上）が2021年に約1.6万人まで膨らんだ。女性は、最大でも2021年の約0.7万人だった（スライドp.11・動画19:14、目測）。
By sex, a difference appears. For men, the net outflow (mostly those aged 45 and over) swelled to about 16,000 in 2021. For women, the largest was about 7,000, also in 2021 (slide 11 at 19:14 in the video, values read by eye).

藤波さんは、地方移住・地方リモートワーク・二地域居住は、中高年男性のライフスタイルに合う、と説明した。そして、地方創生の戦略を組んでいたのはおそらく男性で、男性はそうした動きに目を向けたが、女性には響かなかったのではないか、と推測した。
Fujinami explained that regional migration, regional remote work and dual-location living suit the lifestyle of middle-aged and older men. He then guessed that the people who built the regional revitalization strategy were probably men, and that men turned their eyes to such moves, but it may not have reached women.

若い世代の男女の東京への転入は、同じくらいだ。ただ、男性は大学を出るころに一度地元へ戻る人が多く、45歳以上でも男性は地方へ帰る動きが目立つ。藤波さんは、勝手な推測と断ったうえで、男性は今も家制度を引きずっていて、長男だから、家があるから、親が住んでいるからといった理由で、地元に就職先を求めるのではないか、と話した。女性は一度出ると、家に縛られず、あまり戻ってこないのではないか、とも話した。
Among young people, the inflow to Tokyo is about the same for men and women. But many men return to their hometowns around graduation, and among those aged 45 and over too, men moving back to the regions stands out. Fujinami, saying clearly that it was only his own guess, said men may still be dragging along the old family system, and look for work locally because they are the eldest son, because there is a family home, or because their parents live there. Once women leave, he added, they may not be tied to the household and may rarely come back.

Women, Too, Move for Jobs

藤波さんが言いたかったのは、女性も雇用によって動いている、ということだった。男性が雇用で動くのは当たり前に見えるが、女性も、仕事がどこにあるかより、自分にとってより良い仕事がどこにあるかで動く、と説明した。
What Fujinami wanted to say was that women, too, move because of jobs. That men move for jobs may look obvious, he explained, but women also move by where better work exists for themselves rather than simply where work exists.

これを調べるために、都道府県別の25種類のデータ（女性の労働力率、年収、管理的職業の比率、平均勤続年数、正規雇用比率、大卒人口比、三世代同居比率など）を集め、主成分分析にかけた。第1主成分の軸は、藤波さんの解釈では、右が安定重視、左がキャリア重視だ。右の地域ほど平均勤続年数や三世代同居比率、公務員比率が高く、左の地域ほど女性の賃金や大卒の居住割合が高い。この軸に、2019年の女性の転入超過率（人口千人あたり）を重ねると、右肩下がりの関係が出た（回帰式 y=−0.707x−2.11、決定係数 R²=0.689、スライドp.16・動画28:38）。安定重視の地域ほど女性の転入超過率は低く、キャリア重視の地域ほど高い。ただし、これは都道府県の間の相関で、軸には三世代同居比率のような雇用以外の項目も含まれる。
To look into this, he collected 25 kinds of data by prefecture (the female labor force participation rate, income, the share of managerial jobs, average years of service, the regular-employment ratio, the share of university graduates, the share of three-generation households and so on) and ran a principal component analysis. In Fujinami’s reading, the first axis is stability-oriented on the right and career-oriented on the left. The further right a region is, the higher its average years of service, three-generation household ratio and civil-servant ratio; the further left, the higher women’s wages and the share of university graduates living there. Overlaying the 2019 net inflow rate of women (per thousand residents) on this axis gives a downward-sloping relation (regression y = −0.707x − 2.11, R² = 0.689, slide 16 at 28:38 in the video). The more stability-oriented a region is, the lower women’s net inflow rate; the more career-oriented, the higher. Note that this is a correlation across prefectures, and the axis also includes items other than employment, such as the three-generation household ratio.

勤続年数が長いのは、一見いいことに見える。ただ、藤波さんによれば、言い換えると雇用の流動性が低く、いいポストがなかなか空かない、ということでもある。新卒で社会に出た若い人は、いい位置に就きにくい。東京や大阪では、新しい良い仕事がどんどん出てくる。藤波さんは、労働に関する地域の特性は、人口吸引力そのものではないか、と感じているという。
Long years of service look good at first glance. But, according to Fujinami, put another way it also means low job mobility, so good posts rarely open up. Young people entering the workforce as new graduates find it hard to get a good position. In Tokyo and Osaka, new good jobs keep appearing. Fujinami feels that a region’s labor characteristics may be its pulling power itself.

製造業への依存度も関係している。製造業就業者の比率が高い県ほど、転出超過に占める女性の割合が高い（2020年、決定係数 R²=0.548、2019年に転入超過だった東京圏や愛知県などを除く、スライドp.20・動画39:32）。北関東で顕著だ。製造業は理系・技術系の採用を先行させがちで、結果として男性が多くなり、女性が仕事を見つけにくい。多くの女性は、東京などの情報通信産業へ流れていく、と藤波さんは説明した。
Dependence on manufacturing matters too. The higher a prefecture’s share of manufacturing workers, the higher the share of women among those leaving (2020, R² = 0.548, excluding the Tokyo area, Aichi and others that had a net inflow in 2019; slide 20 at 39:32 in the video). It is especially clear in the northern Kanto area. Manufacturing tends to hire science and engineering people first, which makes the workforce mostly male, so women find it hard to find work. Many women end up flowing to information and communications industries in Tokyo and elsewhere, Fujinami explained.

New Jobs Gathered in the Tokyo Area and Kinki

雇用そのものも、2014〜24年の10年で大きく偏った。2014〜19年は、地方創生1.0の時期で、全国のどの地域でも雇用が増えた。増えた雇用の寄与率は、東京圏が45.1%、近畿が16.0%、東海が11.5%だった。ところがコロナ禍をはさむ2019〜24年は、東京圏が58.0%、近畿が28.6%で、2つで86.6%になる。北陸（▲5.0%）、北海道・東北と中国・四国（ともに▲1.7%）は、雇用が減った（スライドp.17・動画32:44、表の値）。寄与率は、全国の雇用の純増を100としたときの各地域の取り分で、減った地域はマイナスとして差し引かれる。スライドの値は丸めてあるため、8地域の合計はちょうど100にならない。
Employment itself also became heavily skewed over the ten years from 2014 to 2024. In 2014–19, the period of “Regional Revitalization 1.0,” jobs grew in every region of the country. The contribution to the increase was 45.1% for the Tokyo area, 16.0% for Kinki and 11.5% for Tokai. But in 2019–24, which spans the pandemic, it was 58.0% for the Tokyo area and 28.6% for Kinki, which together make 86.6%. Hokuriku (▲5.0%), Hokkaido-Tohoku and Chugoku-Shikoku (both ▲1.7%) lost jobs (slide 17 at 32:44 in the video, values from the table). The contribution is each region’s share when the nation’s net increase in jobs is set to 100, and regions that lost jobs are subtracted as negatives. The slide’s values are rounded, so the eight regions do not add up to exactly 100.

産業別（全国、2019〜24年）に見ると、増えたのは医療・福祉（寄与率65.5%）と情報通信業（47.9%）で、建設業（▲12.6%）、金融・保険（▲8.4%）、製造業と卸売・小売（ともに▲5.0%）は減った。藤波さんは、情報通信業の本社は全国の約6割が東京に集まっている、と話した。
By industry (nationwide, 2019–24), what grew was medical care and welfare (contribution 65.5%) and information and communications (47.9%), while construction (▲12.6%), finance and insurance (▲8.4%), and manufacturing and wholesale and retail (both ▲5.0%) shrank. Fujinami said that about 60% of the nation’s information and communications headquarters are concentrated in Tokyo.

Young Women Went into Information and Communications

女性に絞って見る。25〜34歳の女性（全国）で増えた雇用を100とすると、情報通信業の寄与は85.7だった。東京圏だけで見ても、情報通信業の雇用は、2019〜24年の5年間に、25〜34歳の女性で50.0%、男性で33.3%増えた（スライドp.18・動画35:26、表の値）。ITやエンジニアは女性に就きにくいという印象があるが、藤波さんは、情報通信業は女性を積極的に採用し、採用した女性の中にはプログラマーとして育成していた様子が読み取れる、と話した。なお、この表の寄与率や、情報通信業の伸び率（50.0%、33.3%など）は、万人単位で丸めた増減から計算されたものとみられ（女性25〜34歳・全国の列の値が14分の何かになっている）、小数点以下の細かい精度はない。
Looking at women only. If the growth in jobs for women aged 25–34 nationwide is set to 100, the contribution of the information and communications industry was 85.7. Looking at the Tokyo area alone, jobs in that industry grew 50.0% for women aged 25–34 and 33.3% for men over the five years 2019–24 (slide 18 at 35:26 in the video, values from the table). There is an impression that IT and engineering are hard for women to enter, but Fujinami said the data show the information and communications industry actively hired women and trained some of them to become programmers. Note that the contributions in this table, and the growth rates for the information and communications industry (50.0%, 33.3% and so on), appear to be calculated from changes rounded to units of 10,000 people (the values in the column for women aged 25–34 nationwide are fractions of 14), so the decimals are not precise.

理由の一つとして、藤波さんはジェンダーギャップの小ささを挙げた。産業別に、正規職の賃金・役職の男女差（横軸）と、正規雇用比率の男女差（縦軸）を並べると、日本の多くの産業は右肩下がりの直線の近くに並ぶ。その中で、情報通信業だけが直線から離れ、ギャップの小さい側にいる（スライドp.19・動画37:24）。藤波さんは、情報通信業は伸びている産業なので、もともと社内規定に男女差がなかったり、人を積極的に取りたいので格差を廃止してきたりした、という見方を示した。
As one reason, Fujinami pointed to the small gender gap. If industries are plotted with the gender difference in regular workers’ wages and positions (horizontal axis) against the gender difference in the regular-employment ratio (vertical axis), most Japanese industries line up near a downward-sloping straight line. Among them, only information and communications sits away from the line, on the side with a small gap (slide 19 at 37:24 in the video). In Fujinami’s view, because information and communications is a growing industry, it may have had no gender difference in its internal rules to begin with, or abolished gaps because it wanted to hire people actively.

質疑では、この点を東京への偏りとあわせて問われた。藤波さんは、女性に選ばれているので、放っておけばこうした産業にさらに女性が流れ込むだろう、と答えた。ただし、今回のデータは2024年までで、雇用はコロナ禍のような社会的な出来事で大きく変わる。次の5年のリード産業は分からず、AIの登場で産業構造が変わるかもしれない、とも付け加えた。
In the Q&A, this point was raised together with the skew toward Tokyo. Fujinami answered that because these industries are chosen by women, even more women would flow into them if left alone. However, he added that this data runs only to 2024, and that employment changes a lot with social events like the pandemic. He does not know what the leading industry of the next five years will be, and the arrival of AI may change the industrial structure.

Male Workers Keep Declining; the Growth Is in Women

男女比も動いている。25〜34歳の男女比（男性÷女性）を2000年と2020年で比べると、東京はほぼ1対1に近づき、地方では男性の割合が高まった。藤波さんは、東京は元々働く場だったので男性が多かったが、この20年で男女がほぼ同数になり、結婚のパートナーを見つけやすい地域になった、と説明した。製造業の強い栃木県は男性寄りに（約1.08から約1.16）、かつて女性が多かった大分県は男女ほぼ同数に（約0.92から約1.03）なった（スライドp.21・動画41:06、目測）。結果として、地方では男性の未婚の割合が上がる、と藤波さんは説明した。
The sex ratio is moving too. Comparing the male-to-female ratio of those aged 25–34 between 2000 and 2020, Tokyo came close to one to one, while the share of men rose in the regions. Fujinami explained that Tokyo was originally a place to work, so it had more men, but over these 20 years men and women became about equal, and it turned into a place where it is easy to find a partner to marry. Tochigi, with strong manufacturing, shifted toward men (about 1.08 to about 1.16), and Oita, which once had many more women, became about even (about 0.92 to about 1.03) (slide 21 at 41:06 in the video, values read by eye). As a result, Fujinami explained, the share of unmarried men rises in the regions.

自治体のオンライン婚活には、期待薄だという。女性の仕事が地方になければ、結婚しても、東京で使っていた能力をどこで生かすのかという問題になり、地方へ引っ張ってこられない。
He is not hopeful about municipal online matchmaking. If there are no jobs for women in the regions, then even after marriage the question becomes where a woman could use the skills she used in Tokyo, and she cannot be drawn to the regions.

もう一つ、働き手の構造が変わっている。藤波さんは、15〜64歳の男性の就業者は、1990年以降ずっと減っている、と話した。男性が定年退職した穴を、もう男性では埋められない。増えているのは女性で、働いていなかった人が働き始めている。女性を取り込めない企業は、ますます人を取れなくなる。ただ働かせればいいのではなく、より良い待遇を与える必要がある、と話した。
Another change is in the structure of the workforce. Fujinami said that the number of men aged 15 to 64 in work has kept falling since 1990. The gap left by retiring men can no longer be filled by men. It is women who are increasing, as people who were not working start to work. Companies that cannot bring in women will find it harder and harder to hire. It is not enough to simply have them work; better treatment is needed, he said.

正規と非正規で見ると、2024年4月以降、男女とも非正規が減る月が出てきた。2015〜25年の11年間、ほぼ一貫してプラスだったのは女性の正規雇用だ（2023年に一度だけほぼゼロになった、スライドp.23・動画44:52）。男性の定年退職の穴を女性の非正規で埋めようにも、その人たちも減っていく。正規雇用で待遇を与えないと人を取れない時代が、目前に迫っているという。
By regular and non-regular work, from April 2024 there have been months when non-regular employment fell for both men and women. What has stayed positive almost throughout the 11 years from 2015 to 2025 is regular employment of women (it dropped to nearly zero once, in 2023; slide 23 at 44:52 in the video). Even if companies try to fill the gap of retiring men with non-regular women, those numbers will also decline. He said an era is approaching in which people cannot be hired unless they are given regular employment and good treatment.

Changes Inside the Big Cities

もう一つ、藤波さんが講演で紹介したレポートにある、大都市の中の動きがある。東京都区部と20の政令指定都市を合わせた「21大都市」の転入超過（日本人）は、2024年の123,541人から、2025年の109,016人に、14,525人（11.8%）減った。転入超過であることは変わらないが、勢いは落ちた（スライドp.28・動画53:58、表の値）。
One more thing: there is movement within the big cities, which Fujinami introduced from a report he presented in the talk. The net inflow (Japanese nationals) of the “21 big cities,” meaning Tokyo’s 23 wards and the 20 designated cities together, fell by 14,525 people (11.8%), from 123,541 in 2024 to 109,016 in 2025. It is still a net inflow, but the momentum has dropped (slide 28 at 53:58 in the video, values from the table).

内訳は偏っている。東京圏の都市部は▲13.7%、大阪圏の都市部は▲20.3%と大きく減り、名古屋市は+8.1%と増えた（表の値）。スライドの散布図では、2025年の地価上昇率が5%を超えた東京都区部、千葉市、大阪市で、転入超過が大きく減っている。一方、地価がそれほど上がっていない川崎市や名古屋市などは、転入超過が増えた（目測）。若い人が中心の転入者が大都市に入りづらくなり、周辺の都市に分散している、と藤波さんは見ている。
The breakdown is uneven. The urban parts of the Tokyo area fell a large ▲13.7%, the urban parts of the Osaka area ▲20.3%, while Nagoya rose +8.1% (values from the table). In the slide’s scatter plot, the net inflow dropped sharply in Tokyo’s 23 wards, Chiba and Osaka, where 2025 land prices rose by more than 5%. Meanwhile, cities where land prices have not risen so much, such as Kawasaki and Nagoya, saw their net inflow grow (values read by eye). Fujinami sees newcomers, mostly young people, finding it harder to enter the big cities and spreading out to surrounding cities.

出生数にも動きがある。2025年は、東京都区部が+837人（+1.4%）、千葉市が+0.5%、大阪市が▲0.1%で、藤波さんは、この3つの都市では出生数は底打ちしたと見ている。受け皿になった川崎市（▲2.9%）、神戸市（▲3.9%）、相模原市（▲7.8%）では、大きく減った（2024年と2025年の1〜11月の実績値の比較、スライドp.30・動画57:00、表の値）。藤波さんは、まだ推測の段階だと断ったうえで、東京都区部、千葉市、大阪市には子育てをする人たちが入ってきていて、子どもを持たない人たち（独身を含む）が入りづらくなっているのではないか、と話した。
The number of births has also moved. In 2025, Fujinami sees births as having bottomed out in Tokyo’s 23 wards (+837, +1.4%), Chiba (+0.5%) and Osaka (▲0.1%). They fell sharply in Kawasaki (▲2.9%), Kobe (▲3.9%) and Sagamihara (▲7.8%), which took in the overflow (a comparison of actual figures for January to November 2024 and 2025; slide 30 at 57:00 in the video, values from the table). Fujinami, saying it was still at the stage of a guess, said people raising children may be moving into Tokyo’s 23 wards, Chiba and Osaka, while people without children (including single people) find it harder to move in.

子育て世帯は、住宅補助や公営住宅、子育て向けのアフォーダブル住宅といった支援で、不動産価格の高騰の影響を抑えられる。しかも、子育て世帯の多くは共働きで、職場と家が近くないとやっていけない。保育所に迎えに行くのに片道1時間半かけて通勤するのは、限界がある。だから、なるべく職住近接を目指す、と藤波さんは説明した。
Families raising children can soften the impact of soaring property prices through support such as housing subsidies, public housing and affordable housing for child-rearing. Moreover, many of those households are dual-income, and cannot manage unless home and workplace are close. Commuting an hour and a half one way while also picking a child up from nursery has its limits. So they aim for living near work, Fujinami explained.

そこから、藤波さんは新しい課題を指摘した。都市部の子育て支援の強化は、決して悪いことではない。ただ、不動産価格の上昇と合わさると、子どものいない世帯が都市部から排除される可能性がある。地価が上がって低所得層が追い出されるジェントリフィケーションと同じ構図が、子どもの有無で起きるかもしれない、という。
From there, Fujinami pointed out a new issue. Strengthening child-rearing support in cities is by no means bad. But combined with rising property prices, households without children may be pushed out of the cities. The same structure as gentrification, where land prices rise and low-income residents are driven out, might arise along the line of whether people have children, he said.

Fujinami’s Recommendations

スライド「人口流出への対処」と質疑から、提言を整理する。地方については、次の4つだ。
I organized the recommendations from the slide “Responding to Population Outflow” and the Q&A. For the regions, there are four.

• 目の前の転出超過に焦らない。消滅可能性自治体の議論に踊らされない。移住促進に金を投じても、リターンは小さい。基礎自治体同士の移住者獲得の競争を煽らない。
  Don’t rush over the current net outflow, and don’t be swayed by the debate on “municipalities at risk of disappearing.” Putting money into migration promotion gives small returns. Don’t stir up competition among municipalities to attract migrants.
• 地方の企業に、高度人材（大学や高専を出た人）の採用を促す。地方の企業は、人材獲得競争で東京の企業と同じ土俵にある。東京の企業が地方在住者にオンラインで働いてもらうことは、すでに普通に行われている。
  Encourage regional companies to hire skilled people (those with university or technical college degrees). In the competition for talent, regional companies stand on the same ground as Tokyo companies. Tokyo companies having people who live in the regions work online is already common.
• 男女比の是正に配慮する。女性雇用の質と量を改善し、ジェンダーギャップと、固定的な性別役割分担の意識を改める。
  Pay attention to correcting the sex ratio. Improve both the quality and quantity of women’s employment, and change the gender gap and the awareness of fixed gender roles.
• 自治体のオンライン婚活には、期待しない。結婚が、夫婦2人で働いて家を支える形に変わっているので、女性の仕事が地方になければ、定着は期待できない、という理由だ。
  Don’t expect much from municipal online matchmaking. The reason: marriage has become two people working to support a household, so unless there are jobs for women in the regions, settling there cannot be expected.

大都市については、次のようなことを挙げた。
For big cities, he named the following.

• 雇用を、周辺の都市や郊外に分散させる。若い人にとって魅力的な雇用を、各地域に増やす。
  Disperse jobs to surrounding cities and suburbs, and increase jobs attractive to young people in each area.
• 在宅勤務を復活させる。コロナ禍に導入した企業の多くがやめたが、子育てを考えるとメリットが大きい。国として、もう一度力を入れる。
  Bring back working from home. Many companies that introduced it during the pandemic have dropped it, but the benefits are large when you think about child-rearing. The national government should put effort into it again.
• 現金給付や無償化は、全国一律でやる。住んでいる場所で受けられる金額が違うのは、権利として基本的におかしい。東京の水準に合わせて、全国を引き上げる。
  Make cash payments and fee waivers uniform nationwide. It is fundamentally wrong, as a matter of rights, that the amount you receive depends on where you live. Bring the whole country up to Tokyo’s level.
• 医療などのインフラの水準も合わせる。不妊治療のクリニックは東京に集中し、無痛分娩は地方ではほとんど行われていない、という。
  Match the level of infrastructure such as healthcare. Fertility clinics are concentrated in Tokyo, and epidural childbirth is hardly offered in the regions, he said.
• 不動産価格の安定は重要だが、民間の取引なので、国にできることは限られる。金利を上げれば冷えるが、そのためだけにはやれない（質疑）。
  Stable property prices matter, but it is private trading, so there is a limit to what the national government can do. Raising interest rates would cool it, but that cannot be done for that purpose alone (Q&A).

From the Q&A: Regional Companies Becoming Healthy

質疑で、地方創生とは結局何か、と問われた藤波さんの答えは、地方の企業が元気になること、それ以外にはない、というものだった。地方の企業が雇用を失いつつある状況では、地方の活性化といっても限界がある。多くの地方の企業は中小企業が主役なので、そこが生産性を上げ、高い給料で人を採用していく状況をつくる以外にない。
Asked in the Q&A what regional revitalization is in the end, Fujinami answered that it is regional companies becoming healthy, and nothing else. When regional companies are losing jobs, there is a limit to revitalizing a region. Because small and medium-sized companies are the main players in many regions, there is no way other than creating a situation in which they raise productivity and hire people at high pay.

国の政策では、2026年6月時点の政権が、地方創生の発展形として「地域未来戦略」を打ち出した。雇用をしっかり育てる方向を前面に出した点を、藤波さんは必然だと見ている。ただ、その中心のツールの「クラスター戦略」には期待していない。2000年ごろにも国が取り組み、10〜20個ほどあったクラスターが、ほとんど残っていない。産業は国の思うようにはならず、芽のない地域に特定の産業の種をまくのは厳しい。すでに核になる産業がある地域で、周辺の企業を支援してクラスターにするのが、本来の姿だと藤波さんは見ている。
In national policy, the government at the time of the talk (June 2026) put forward the “Regional Future Strategy” as a development of regional revitalization. Fujinami sees it as inevitable that it put the direction of building up jobs to the front. But he does not expect much from its central tool, the “cluster strategy.” The national government tried it around 2000 too, and of the roughly 10 to 20 clusters that existed, almost none remain. Industries do not grow the way the state wants, and it is hard to sow the seed of a particular industry in a region with no prospects. He believes the proper form is to support the surrounding companies in a region that already has a core industry, and turn that into a cluster.

経済産業省の取り組みとして、中堅・中小企業の社長に売上100億円を目指すと宣言してもらい、計画づくりや体質改善を国が支援する制度が、2025年から始まっている。講演の時点で、約3,000社が手を挙げているという。藤波さんは、地方の中小企業が自ら成長していかなければ若い人は定着できないとして、これを良い流れだと評価した。
As an initiative of the Ministry of Economy, Trade and Industry, a program began in 2025 in which presidents of mid-sized and small companies declare they will aim for 10 billion yen in sales, and the national government supports their planning and improvement of company health. At the time of the talk, about 3,000 companies had raised their hands. Fujinami called this a good trend, since young people cannot settle unless regional small and medium-sized companies grow on their own.

事例として、藤波さんは参考資料で、製造業の中小企業を1社紹介した。約150名の醸造機械のメーカーで、20年ほど前まで男性ばかりだった会社が、20年の間に、女性が活躍する会社に変わったという。女性の働きやすさを改善し、女性の昇進の道筋をつくったことが大きい。藤波さんは、中小企業は、社長の意識改革次第で、会社の体質が変わりうる、と話した。
As an example, Fujinami introduced one small manufacturing company in the reference materials. It is a maker of brewing machinery with about 150 employees, and a company that was all men until about 20 years ago changed, within 20 years, into one where women are active. Improving how easy it is for women to work and creating a path for their promotion made the difference. Fujinami said that in small and medium-sized companies, the character of the company can change depending on the president’s change of mind.

質疑の最後に、種鶏の例が出た。藤波さんの説明では、日本で食べられる卵のほとんどは、種鶏（親鶏のもとになる鶏）を海外から輸入した鶏が産んだものだ。昔は日本にも種鶏の産業があったが、1970〜80年代に、国際競争に負けて淘汰された。欧米の数社が、世界中の鳥の遺伝子を集めて品種改良を進め、世界へ届ける産業になったからだ。ただ、日本は世界で唯一、生卵を食べる文化があるといい、ウイルスフリーの安全性や、ご飯に割ったとき形がいい卵といった、日本向けの品種改良もされている。藤波さんは、地方の企業も、世界と戦っていることを意識して品種改良や商品改良をやらなければ、負けてしまう、と締めた。
At the very end of the Q&A, the example of breeding stock came up. According to Fujinami, most of the eggs eaten in Japan are laid by hens whose breeding stock was imported from abroad. Japan once had its own breeding-stock industry, but it lost in international competition and was weeded out in the 1970s and 80s. A few companies in Europe and the United States collected the genes of birds from around the world, advanced breeding, and became an industry delivering to the world. Still, Japan is said to be the only country in the world with a culture of eating raw eggs, and breeds are also improved for Japan, such as for virus-free safety and eggs that hold a good shape when cracked over rice. Fujinami closed by saying that regional companies, too, will lose unless they are conscious that they are competing with the world and keep improving their varieties and products.

My Thoughts

• 若い女性の雇用の伸びの8割以上を、情報通信業が占めていた。ああ、私はこの大群の一部なのかと思った。良い悪いではなく、私も統計の一部なのだと。
  Information and communications accounted for more than 80% of the job growth for young women. I thought, oh, I am part of this crowd. Not in a good or bad way: I am part of the statistics too.
• 東京圏から地方へ出ていったのが主に45歳以上の男性だったというのも、納得した。その年齢ならお金もあるし、男性は女性よりコミュニティに帰属する傾向が弱いのだろう（偏見）。
  It also made sense to me that the people who left the Tokyo area for the regions were mainly men aged 45 and over. At that age they have money, and men probably tend to belong to communities less than women do (my bias).
• 藤波さんのような誠実で丁寧で正直な方々の研究がより真っ当に評価されるようになってほしい。メディアがそんなことをするとは期待できないけど…。とりあえずこれがYouTubeで無料で見れることに感謝
  I hope the research of sincere, careful, and honest people like Fujinami will be valued more fairly. I can’t expect the media to do that, though… For now, I’m just grateful that this can be watched on YouTube for free.

Sources

動画：「人口減少時代を生きる」(12) 藤波匠・日本総研主席研究員 2026.6.8（日本記者クラブ、YouTube）
Video: “Living in an Age of Population Decline” (12) Takumi Fujinami, Principal Researcher, the Japan Research Institute, 2026.6.8 (Japan National Press Club, YouTube)

会見リポート：日本記者クラブ「会見リポート」
Press conference report: Japan National Press Club, “Press Conference Report”

図は、講演スライドの数値（表の値と、グラフからの目測）をもとに、私が作り直した。各図の下に、元のスライドの番号と、統計の出所を書いた。
The charts were redrawn by me from the numbers in the lecture slides (values from tables, and values read by eye from charts). Below each chart, I noted the original slide number and the source of the statistics.

今回のリデザインで、「シニフィエ：私＝シニフィアン：ロゴ」をどう表すかをいじり続ける日々に、せめて数年？数ヶ月？は区切りがつくといいなと思っている。
I hope this redesign finally puts my endless tinkering on pause, for a few years at least (or maybe just a few months?): the tinkering over how to represent “signified : me = signifier : logo.”

しばらくMediumでは執筆していなかった。最後は3月のFigmaのウィジェットを作った話だ。
I haven’t written on Medium in a while. The last one was back in March, about building a Markdown widget for Figma.

今のSNSはもう「ソーシャル」ではないと思っている。個人と個人が対等にやり取りする、パーソナルで親密性の高い場ではなくなった。ビジュアルやタイトルで興味を引き、とにかく発信すること自体が目的化していて、実質的にはマジョリティのメディアが支配する世界になっている。
I don’t think “social media” is social anymore. It used to be a place where individuals talked to individuals, on equal footing, with real intimacy. Now it runs on visuals and titles built to grab attention, where putting something out there has become the point in itself. In practice, it’s become a world dominated by majority media.

Mediumはまだ個人にも人権がある気がするので、長い文章でありのままの考えをだらだら溜める場所として、しばらくここでは記事を書いていなかったけど細々とやっていきたいと思う。これまでの技術やデザインの話に加えて、noteに書いてきた思考整理・読書・哲学系の文章も、これからはここに置いていく。戻ってきて1本目の今回は、私が使っているロゴの話から始めたい。
Medium still feels like a place where an individual has some standing, so I want to keep writing here, quietly, as a place to let my unfiltered thoughts pile up in long, rambling writing, even though I haven’t posted here in a while. Alongside the tech and design posts I’ve written before, the essays I’ve been writing on note, a Japanese blogging platform, mostly about thinking things through, reading, and philosophy, will live here from now on. This first post back starts with the logo I use.

きっかけは、今までポートフォリオサイトに置いていたロゴが自分の中でしっくりきていなかったことである。
It started because the logo I’d had on my portfolio site never quite sat right with me.

合わせて、松岡正剛の『日本文化の核心：「ジャパン・スタイル」を読み解く』をAudibleで聴いたり、海野弘の『日本の装飾と文様』を読んだり、日本文化や芸術についてのYouTubeやポッドキャストに触れる機会が増えていた。
Around the same time, I’d been listening to Seigo Matsuoka’s The Essence of Japanese Culture on Audible, reading Hiroshi Unno’s Traditional Japanese Decorations and Patterns, and taking in more YouTube videos and podcasts about Japanese culture and art.

このような情報、自分の頭の中に入れることで、日本に対する見方が解像度が高くなった、気がする。日本語と言う言語、建築、絵画、人々のコミュニケーション、美意識、道徳、流行語、CM、などなど…。
Taking all of this in, I feel my view of Japan has gotten sharper, as if in higher resolution: the Japanese language, architecture, painting, the way people communicate, the sense of beauty, morals, buzzwords, TV commercials, and so on…

日本は独特であり（日本礼賛を強調したい訳ではない）、時代によって様々な人格みたいなものがある。今生き残っている歴史もほんの1部でしかない。そして「日本人は〇〇だ」とか言われるような、その言い切ってしまう態度にも一歩引いてみることができるようになった。本物の時代を超える教えや考え方などもある一方で、改ざん・希薄化されたものもある。
Japan is distinctive (I’m not trying to sing its praises), and it seems to have had different personalities in different eras. The history that survives today is only a small part of it. I’ve also become able to step back from sweeping claims like “Japanese people are such-and-such.” There are genuine teachings and ways of thinking that outlast their era, but there are also things that have been altered or watered down.

それには、裏にそれを広めたい勢力がいたり、大衆が即座に理解しやすい・合意形成がしやすいからと言う背景が見え隠れする。特に時代が進むについて増えているように思う。
Behind that, I can sometimes glimpse forces that wanted to spread those ideas, or the simple fact that such versions are easier for the public to grasp at once and agree on. And the closer we get to the present, the more often I feel the facts get bent for commercial reasons.

その中で、自分は何を美しいと感じ何を自分の核にしたいのかを考えるようになった。グローバルなデジタルの世界で自分固有のものを出すには、自分が生まれたこの日本について、もっと知見を深めたいと思った。何千年にもわたる、発酵しまくってきた日本の文化からインスピレーションを借りて、何かしらの形にしてみたかった。
Somewhere in that, I started thinking about what I find beautiful, and what I want at my own core. In a global, digital world, I felt that building anything distinctly mine meant going deeper into Japan, the place I was born into. I wanted to borrow from a culture that has been fermenting for thousands of years and turn it into some kind of output.

情けない言い方をすると、すがってみた、となる。ただ私はこれを全く悪いとか卑怯だとかは思っていない。クリエイティビティは既存のものの真似か新結合だからである。ありがたく思いっきり真似させていただく。
Put less flatteringly, I clung to it. But I don’t think that’s bad or cheating at all, because creativity is either imitating what already exists or making new combinations of it. So I’ll gratefully imitate as much as I can.

ポートフォリオサイトのロゴも、今回作ったものに差し替えた。
I’ve now swapped the portfolio site’s logo for this one too.

このロゴは、ひとつの理由から生まれたわけではない。苗字に共通する部首、名前の書体の曲線、最近読んでいる本の装丁、ローマ字にしたときの頭文字。もともと別々に頭の中にあった連想が、たまたま同じ丸と棒の形に重なっただけである。
This logo didn’t come from a single reason. A radical shared by two surnames. The cover of a book I’m reading. The curve of my given name in hand-brushed hiragana. The Roman letter it turns into. A few unrelated associations that happened to have been sitting in my head at once, and happened to land on the same shape.

上の図は、この連想がどう一つの形に収束したかを描いたものである。
The diagram at the top traces how those associations converged into one shape.

一つ目は苗字である。小学校に上がる頃両親が離婚し、私は母子家庭で育った。そのとき、苗字は「浜田（Hamada）」から今の「河野（Kawano）」に変わった。
The first is my surname. My parents divorced around the time I started elementary school, and I was raised by my mother. My surname changed then, from 浜田 (Hamada) to my current 河野 (Kawano).

浜田も河野も、氵（さんずい）を持っている。苗字は変わっても、水に関わる部首だけは変わらずそこに残っていた。氵はもともと、水が流れる様子を描いた象形文字で、線が徐々に崩れて今の三本の点のような形になっている。
Both 浜田 and 河野 carry 氵, the three-dot water radical. The surname changed, but the radical stayed. 氵 began as a pictograph of flowing water; over centuries the lines simplified into the three strokes it is today.

二つ目は、さっき触れた海野弘の本の装丁である。表紙は上に丸、下に長方形という、非常にシンプルな構成になっている。清潔で、まっすぐで、それでいて品がある。
The second is the cover of that Hiroshi Unno book I mentioned above. It’s just a circle above a rectangle, nothing more. Clean, upright, and still dignified.

三つ目は、名前の書体である。私の名前「いおり」は、漢字ではなくひらがなで書く。「いおり」という音は漢字表記の方がよく聞くので、日本人としては少し珍しい部類だと思う。ひらがなにしたのは、母が「ひらがなの方が可愛い」と思ったという、わりと単純な理由による。笑
The third is the shape of my given name. My given name, いおり (Iori), is written in hiragana, not kanji. The sound “Iori” is more commonly seen written in kanji, so writing it in hiragana is a little unusual for a Japanese name. The reason is fairly simple: my mother thought hiragana looked cuter. lol

子どもの頃は漢字の方がかっこいいと思っていた時期もあったが、ひらがなの成り立ちを知った今は、ひらがなで良かったと思っている。そのひらがなつながりで参考にしたのが、紀貫之が935年頃に書いた『土佐日記』、日本最古級の仮名日記文学である。教科書的なひらがなではなく、筆で書いたときの流れるような崩し方で「いおり」を捉えたとき、丸みのある曲線が頭に残った。
There was a stretch as a kid when I thought kanji looked cooler, but now that I know more about the history of hiragana, I’m glad it’s hiragana. Because of that hiragana connection, I borrowed a curve from the Tosa Nikki, one of the earliest works of Japanese diary literature, written by Ki no Tsurayuki around 935. When I picture いおり not in textbook hiragana but in the loose, brushed form that diary was written in, what stays with me is the roundness of the curve.

図に載せているのは935年の原本ではない。本来は原本を参照するのが望ましいのは承知しているが、良質な画像資料として入手できたのが早稲田大学図書館が所蔵する1707年の版本だったため、こちらを使っている。「い」「お」はこのページから、「り」はこの本文ページから採っている。
The page in the diagram isn’t that 935 original. Ideally the original would be the more authoritative source, but a 1707 edition held by Waseda University Library was what I could find as good-quality image material, so that’s what I used: い and お come from this page, り from this page of the main text.

苗字の部首、本の装丁、名前の曲線。この三つが別々に頭の中にあって、それが丸を二つと棒一本という同じ形に重なった。
A radical in my surname, a book cover, the curve of my given name. Three separate things sitting in my head, and they converged on the same shape: two circles and a bar.

さらに、この形はローマ字にしたときの名前「Iori」の頭文字「i」にも見える。iを人のシルエットとして見ると、上の丸が頭（脳）、下の棒が体になる。アイコンをデザインするときは、正方形のキャンバスの中に収める形で考えるのが基本である。その正方形の中心に来るのが、頭の丸、つまり脳にあたる部分になる。認知も感覚も、何かを生み出すことも、すべての起点は脳にある。体も、脳が考えることそのものも、その周りにあるものだと思っている。
On top of that, the shape also reads as a lowercase “i,” the first letter of “Iori” in Roman characters. Read that “i” as a human silhouette and the top circle becomes a head, a brain, and the bar below becomes a body. Icon design usually works within a square canvas, and the head, the brain, is what sits at the center of that canvas. I think of the brain as the origin of all cognition, all sensation, everything a person makes. The body, and the thought the brain produces, sit around that center.

自分を三つの言葉で表すなら、Analytical, Inquisitive, Associativeだと最近気づいた。以前、海外で働く日本人のインタビューポッドキャストを聞いていて、同じ質問をされたら自分は何と答えるだろうと考えたのがきっかけである。今回のロゴ作りは、そのうちのAssociativeな部分がそのまま出てきたような感覚だった。別々にあった連想を、後から一つの形につなげる。自分の頭は、常にそうという訳でないが普段からそういう働き方をしているのだと思う。
If I had to describe myself in three words, I’d say Analytical, Inquisitive, Associative. I landed on those recently, while thinking back to a podcast interview with a Japanese person working abroad. They were asked the same three-word question, and I found myself asking what I would have said. Making this logo felt like the Associative side answering on its own: taking things that were sitting apart and pulling them into one shape afterward. Not always, but I think that’s how my head usually runs.

次からは、哲学的なことも含め、自分がそのとき考えていることをありのままに書いていくつもりである。
From the next one on, I plan to write down whatever I’m thinking about as it is, philosophical questions included, but not limited to them.

How I described my ideal Figma widget to Cursor, and it handled everything from architecture to edge cases

I write a lot of Markdown in my daily work. Notes, specs, design token tables. When I started putting those into Figma, I looked for widgets that could render Markdown on the canvas. I found a few. They handled headings and bullet points fine.

Then I tried a table.

The widget displayed | Header | Header | and |---|---| as plain text. The pipe characters, the dashes, all of it, just sitting there on the canvas like raw code. I tried three different widgets. Same result every time.

I didn’t want to build a Figma widget from scratch. But I also didn’t want to keep staring at raw pipe characters on my canvas. So I opened Cursor and described what I wanted.

What I Asked For

I gave Cursor one prompt:

I want a Figma Widget that renders Markdown on the canvas. Existing widgets don’t render tables properly. I want tables to display as visual tables with headers, borders, and alternating row colors. The width should be adjustable.

No technical specification. No architecture diagram. Just what I wanted as a user. The AI chose TypeScript and esbuild, picked a Markdown parsing library, and started building.

What came back was a widget that renders headings, lists, code blocks, blockquotes, checkboxes, and clickable links.

The feature I cared about most, visual tables, worked as I’d described. Header rows get a background color. Data rows alternate between light and dark. Cells have borders. Long text wraps. Inline formatting works inside cells too.

What the Widget Does

Perfect Markdown — available on Figma Community

You place the widget on your canvas, click Edit, paste your Markdown, and hit Save. Four width presets (400, 600, 800, 1200px) cycle with a toolbar button. Light and dark themes switch from a dropdown.

Another feature that turned out to be useful: color swatches. If you write a hex color code like #1a4cb3 anywhere in your text or table, the widget renders a small colored rectangle next to it. Handy for design token documentation.

The edit workflow is simple. Click Edit on the toolbar, a modal opens with a textarea, paste or type your Markdown, click Save.

When the First Approach Broke

I loaded the widget into Figma. Error: Unexpected token .... Figma's Widget runtime is ES5 only; marked.js uses modern JavaScript and wouldn't run there.

I pasted the error into Cursor. The AI tracked it down, saw the library couldn’t run in that environment, and proposed a custom parser in ES5-compatible code. I hadn’t asked for that and didn’t know ES5 was the limit. It built the parser. Roughly 300 lines. It works.

The Conversation That Built the Widget

What I find interesting about this project is how little I specified. The conversation with Cursor went roughly like this:

1. “Build a Markdown widget for Figma with visual tables” → Initial version built
2. “It doesn’t work, here’s the error” → ES5 constraint discovered, custom parser built
3. “Add checkboxes and clickable links” → Parser extended in one pass
4. “Add light/dark themes and width presets” → UI controls added

Each step was a request in plain language. I described what I wanted as a user, not as a developer. The AI made the technical decisions: which API methods to use, how to structure the parser, how to calculate table cell widths, how to render alternating row colors.

This was my third Figma plugin built with Cursor. With my first (PPTX to Figma), I learned that the AI could investigate bugs, not just write code. With my second (Bulk Screenshot Importer), I learned it could invent algorithms I hadn’t thought of. With this one, I learned that I could describe a complete product in a few sentences and get something that works.

What Still Falls Short

The widget isn’t perfect. Images render as placeholders because Figma’s Widget API restricts external image loading. Niche Markdown features like footnotes and definition lists aren’t supported. The source code ended up as a single large file rather than the modular structure I’d prefer.

These are real limitations. But for the use case I built it for, notes and specs and token tables on a Figma canvas, it covers what I need daily.

What This Project Taught Me

Three plugins in, the pattern is clear. I describe what I want. The AI builds it. Something breaks. The AI fixes it. I ask for more features. The AI adds them.

The part that keeps getting better is the conversation. With each project, I’ve gotten more comfortable describing outcomes instead of implementations.

“I want tables that look like tables” instead of “parse pipe-delimited rows and render them as AutoLayout frames with alternating background fills.” The AI translates the intent into code. And when the code hits a wall, like an ES5 runtime nobody warned me about, the AI figures out the workaround too.

I’m not sure where the ceiling is for this workflow. But three plugins in, I haven’t hit it yet.

Try It

Perfect Markdown — Figma Community

If you write Markdown in Figma for notes, specs, or documentation, and you’ve been frustrated by tables showing up as raw text, give this a try. Place the widget, paste your Markdown, and see if it covers what you need.

I’d welcome feedback on what works and what doesn’t. Every Markdown document is a little different, and real-world usage is the best way to find what the parser misses.

Content Declaration

-Author: Iori Kawano
-AI Assistance: Yes — AI (Cursor) used for development and editing. Original draft by author.
-Writing Process: Human-directed, AI-drafted
-Date: 2026–03–17

Tags: Figma, Markdown, Design, AI, Vibe Coding

I’ve been collecting hints about the future. Not systematically. More like picking up interesting rocks on a beach walk.

A podcast episode here. A blog post there. A defense company’s internal strategy document. A historian’s bleak-yet-hopeful take on human nature. Each came from a completely different field. But over the past few months, I noticed something: they kept pointing in similar directions.

This article is my attempt to lay those clues on the table and see what patterns emerge. Some sections are more developed than others. This is thinking in progress, not a finished thesis.

When Value Shifts to Judgment

Kazuto Ataka is one of Japan’s most influential thinkers on AI and strategy. He’s the Chief Strategy Officer at Z Holdings (the parent company of Yahoo Japan and LINE), former McKinsey consultant, and neuroscientist by training. His blog and YouTube channel are required reading for anyone in the Japanese tech scene.

Ataka describes a shift in where economic value lives, which I’d summarize as:

In the manufacturing era, value was in physical goods.

In the IT era, value moved to information.

In the AI era, value moves to judgment.

He frames this through what’s known as the “smile curve” in manufacturing economics. Traditionally, value concentrates at both ends of the value chain: upstream (R&D, design) and downstream (branding, customer service), with manufacturing in the middle adding less value.

Ataka argues that AI flattens this curve into what he calls a “ko shape” (コの字型, after the Japanese character コ). The middle gets automated entirely. What remains for humans is direction-setting at the top and final quality judgment at the bottom.

Ataka writes (my translation from Japanese):

What humans need in an era where AI succeeds too well is not advanced technical skill, but “will” (意志), a kind of perception born from raw, lived experience, close to wild instinct.

What do you accept and what do you refuse? Where should you head and where shouldn’t you?

The ability to rewrite your own “optimization function” itself. That may be the core of our survival strategy going forward.

What struck me was that phrase: “rewrite your own optimization function.” In an era where AI handles optimization, the critical human skill becomes deciding what to optimize for. Not a technical skill. Something closer to instinct, shaped by lived experience.

References:

• Kazuto Ataka’s blog (Japanese)

A 1970 Prediction That’s Looking Eerily Accurate

In 1970, Kazuma Tateishi, the founder of Omron (the Japanese automation and electronics company), presented a future prediction model called SINIC Theory at an international conference. SINIC stands for “Seed-Innovation to Need-Impetus Cyclic Evolution.” It maps out the trajectory of human civilization in stages.

SINIC Theory: A Compass for the Future

What’s remarkable is that its predictions for the early 21st century have been surprisingly on track. Here’s the relevant timeline:

Optimization Society (2005–2024): Information technology optimizes systems at scale

• Autonomous Society (2025–2033): Individuals pursue spiritual and inner fulfillment, supported by autonomous systems
• Natural Society (2033 onward): Humans, technology, and nature form an integrated whole

The “Natural Society” is described as a kind of hyper-primitive state. Not a regression, but an advanced stage where technology becomes so sophisticated it disappears into daily life. Control gives way to organic harmony. People find joy in being part of the ecosystem rather than dominating it.

This raises a question I keep coming back to:

Does a utopian transition require a catastrophic event first?

History suggests that paradigm shifts are often preceded by destruction. Feudal society collapsed before the Industrial Revolution took hold. What happens before 2033?

References:

• Omron official: SINIC Theory
• Dentsu Research: “The society predicted in 1970 through 2033” (Japanese)

Technology Meets Geopolitics

Makoto Shiono is a Japanese strategist who has consistently written about the intersection of technology and international politics. He currently leads the emerging technology group at the Asia Pacific Initiative’s Institute of Geoeconomics, and serves as Managing Director at Industrial Growth Platform, Inc. (IGPI).

Two of his books stand out:

• Digital Technology and the Dynamics of International Politics (2020), analyzing how technology shifts national power through a geoeconomic lens
• Telecommunications Geopolitics 2030 (2022), examining internet fragmentation and the geopolitical fractures of 5G

Looking back from 2026, with the Ukraine invasion (2022) and ongoing Middle Eastern conflicts as context, Shiono’s 2020 analysis of technology-geopolitics convergence feels prescient.

Anduril’s “China 27” Strategy

My vague feeling that “the biggest wave hasn’t hit yet” aligns with concrete moves by Anduril, the U.S. defense technology company founded by Palmer Luckey (of Oculus Rift fame).

• Anduril’s internal strategy, reportedly called “China 27,” assumes a possible Chinese invasion of Taiwan by 2027 and plans to have key defense projects operational by then (Business Insider)
• They’ve opened an office in Taipei and signed a memorandum with Taiwan’s National Chung-Shan Institute of Science and Technology (Anduril)
• Their approach focuses on swarms of small, affordable platforms as “technological deterrence” to prevent armed conflict in the first place (Grand Fleet)

Connecting This to SINIC Theory

SINIC Theory predicts the Natural Society begins around 2033. Anduril is planning around 2027. If major geopolitical upheaval occurs in that window, the optimistic future that SINIC describes might only be reachable after passing through significant turmoil.

I want to be clear: I don’t have conviction that a Taiwan crisis is inevitable. It’s more of a vague impression. But the timeline overlap between SINIC’s prediction and defense industry planning is hard to ignore.

A Society That Rewards Good Behavior

Shifting from geopolitics to social design. Jad Tarifi, co-founder and CEO of Integral AI (a Tokyo-based company), caught my attention. He’s a former Google engineer who helped build Google’s first generative AI team.

Tarifi advocates for what he calls the “Alignment Economy,” a system designed to maximize human freedom (Singularity Weblog / AI and the Future of Work Podcast #350).

From what I remember of his talks (this is my interpretation, not his exact words), the vision goes something like this:

• Go to the gym? You get rewarded.
• Eat healthy? Rewarded.
• Help someone? Rewarded.

Rewards meaning something like points or credits. Good behavior becomes tangibly incentivized.

My first reaction was pure excitement.

I’d love to live in that world. I’d go to the gym every day!!!!!

But my second reaction was skepticism: if everyone does good things for points, is it still goodness? Point-motivated kindness isn’t exactly kindness.

I didn’t have a solution to this tension. But I had an intuition: maybe complex system design could resolve it. Not simple “action = points,” but something more nuanced. Something like what I discovered next.

Complex Systems Over Simple Rules

Audrey Tang, Taiwan’s first Digital Minister (appointed 2016), offers a concrete example of what I mean by “complex systems helping humans.”

Tang’s core belief is that social division stems not from disagreement, but from lack of mutual attention. Her approach: don’t suppress conflict. Transform it into co-creation. She built platforms like vTaiwan for citizen participation and government transparency.

Quadratic Voting

The mechanism that fascinated me most is Quadratic Voting, developed by economist Glen Weyl (Microsoft Research) and adopted in Taiwan.

In standard voting, everyone gets one vote. In Quadratic Voting:

• Each person receives 99 points
• Casting 1 vote costs 1 point
• Casting 2 votes on the same issue costs 4 points (2²)
• Casting 3 votes costs 9 points ( 3²)
• And so on

The cost of expressing strong preference grows quadratically. This means you can concentrate your votes on issues you care deeply about, but it gets expensive fast. The result: minority opinions get more room, and tyranny of the majority is structurally dampened.

Why Counter-Intuitive Systems Might Be Better

This system is not intuitive. It’s hard to explain at a dinner party.

But maybe that’s the point. In a world as complex as ours, maybe intuitive systems (one person, one vote, majority wins) are actually the dangerous ones. Simple majority rule can easily produce majority tyranny. Like complex systems in nature, counter-intuitive mechanisms might produce fairer, more sustainable outcomes.

Could the “reward society” problem (goodness-for-points isn’t real goodness) be solved by similar complexity? Not simple “good deed = credit,” but weighted systems that account for context, ripple effects, and behavioral patterns?

Honestly, I can’t propose a specific solution. I don’t have the knowledge or intellect to push this question much further right now. But I hold onto the intuition: something exists in the direction of “complex, non-intuitive systems that somehow work things out.”

When Technology Becomes Invisible

A quick thought experiment that occurred to me one afternoon.

I’m a Spotify user who wears earphones most of the day, switching between podcasts and music. When I’m doing deep work at my computer, I switch to focus-friendly BGM. When I go for a walk and my brain has spare capacity, I wish Spotify would automatically suggest a podcast or audiobook.

As an Apple Watch user, I’d love to approve or reject these suggestions with a hand gesture. Frictionless.

What kind of algorithm would this need? Body data (posture, heart rate, brain activity), behavioral data (what I’ve been reading, watching, doing across all devices), environmental data (weather, time of day, location).

It sounds amazing. I genuinely want this experience.

But flip the perspective: a computer would be choosing what enters my ears, and by extension, shaping my thoughts. The “technology melting into daily life” that SINIC Theory envisions sits on a razor’s edge between convenience and dystopia.

Keita Watanabe, a Japanese interaction designer, coined the concept of Tokeru Design (“melting design”) in his 2015 book. He describes technology dissolving into life so completely that you stop noticing it, like how you don’t feel your clothes after wearing them for a few minutes. Beautiful in theory. Unsettling when applied to cognition.

Mutual Aid by Process of Elimination

Ryunosuke Fukai is the CEO of COTEN, a Japanese company that has built a database spanning 3,500 years of human history. Their podcast, COTEN RADIO, is one of the most popular in Japan, with a mission to provide “opportunities for meta-cognition” through deep historical analysis.

In a recent podcast episode (Spotify: “Inside Vision,” Feb 7, 2026, ~1h41m, Japanese), Fukai spoke about his personal philosophy with striking honesty.

He described himself as someone who finds living tedious, who doesn’t feel happy when others are happy, who rates his own “goodness score” below average. “I’m probably closer to a sociopath,” he said matter-of-factly.

But here’s his reasoning: no matter how much he denies his own goodness, he cannot deny that he is Homo sapiens. And Homo sapiens, as a species, has one undeniable characteristic that separates us from every other animal: we cooperate at massive scale. No other species builds social security systems. No other species runs joint projects involving tens of millions of individuals.

After eliminating everything else, denying goodness, denying idealism, denying every possible justification, the one thing left standing is: “I am Homo sapiens, and Homo sapiens are fundamentally mutual-aid animals.”

“It’s either die, or help each other. There’s nothing else to do.”

This isn’t idealism, Fukai emphasizes. It’s a conclusion reached by process of elimination, through deep study of philosophy, social science, and history. And to those who say mutual aid can’t work in a market economy, he responds: “Mutual aid is the fundamental nature of Homo sapiens, and it has been making incremental progress throughout history. What is the rational basis for claiming it stops right now?”

What I Took Away

I found this deeply compelling. Who am I to resist the nature of my own species? If a historian who has studied 3,500 years of human behavior concludes that mutual aid is one of our core behavioral principles, arrived at through elimination of all alternatives, I’m inclined to follow that instinct.

Looking back, I think my interest in reward societies and complex systems also stems from this: an intuition that systems aligned with our mutual-aid nature should work better than those fighting against it.

Everything Connects (Somehow)

These clues came from completely different fields: AI strategy, a 1970 prediction model, defense technology, behavioral economics, digital democracy, interaction design, and history.

But they keep overlapping:

• SINIC Theory says we’re heading toward a “natural” society by 2033
• Ataka says the human role in AI age is setting direction and making judgment calls
• Shiono and Anduril suggest geopolitical turbulence may precede that transition
• Tarifi envisions reward systems for good behavior
• Tang and Weyl show that complex, non-intuitive systems can produce fairer outcomes
• Watanabe imagines technology dissolving into life
• Fukai argues mutual aid is our species’ non-negotiable default

I don’t have a unified theory. I’m not sure these clues point to one clear future. But they all seem to orbit the same set of questions: What kind of systems should we build? How do we design for human nature rather than against it? And what will it cost to get there?

I’ll keep collecting clues.

This article was originally published in Japanese on note. This English version includes additional context for international readers.

Content Declaration

Author: Iori Kawano

AI Assistance: Yes. AI used for structuring and editing. Original research, ideas, and voice memos by author.

Writing Process: Human-directed, AI-drafted

Date: 2026–06–08

Tags: Future AI Geopolitics Philosophy Japan

How researching 25+ sites and hand-coloring layers taught me what AI can’t do in cultural design

I started this project with a confident assumption: AI could handle most of the heavy lifting.

I was building an interactive website about ukiyoe, Japanese woodblock prints.

浮世絵タイムライン | Ukiyo-e Timeline

One of the most iconic ukiyoe prints: Katsushika Hokusai, Under the Wave off Kanagawa (c. 1830–32). This full-color nishiki-e was made by layering pigments from multiple carved woodblocks.

The concept was straightforward: show how printing techniques evolved from monochrome ink to full-color masterpieces, using scroll-driven animations where color layers stack one by one, just like the original printing process.

With AI handling code generation and layout exploration, I estimated 2–3 weeks to a working prototype. Three months later, I was still hand-coloring layers in Adobe Fresco. Here’s what happened.

What Ukiyoe Actually Is

For those unfamiliar: ukiyoe (浮世絵) literally means “pictures of the floating world.” These woodblock prints were the mass media of Edo-period Japan (1603–1868). Not precious gallery art, but affordable prints that ordinary people bought and displayed.

The technique evolved dramatically over roughly two centuries:

• Sumizuri-e (墨摺絵): Monochrome prints using only black ink
• Benizuri-e (紅摺絵): Two-to-three-color printing, adding red and green
• Nishiki-e (錦絵): Full-color prints using 10+ separate woodblocks

Each color required its own carved woodblock, applied in a specific sequence. This layering process, one color at a time and precisely aligned, was what I wanted to visualize on the web.

A note on historical accuracy: the actual evolution of ukiyoe printing is more complex than the three-stage model above. There were intermediate forms like tan-e (丹絵) and urushi-e (漆絵). I chose to simplify for clarity, and noted this in the site itself.

References that shaped my understanding:

• Yoshio Kato, “The Evolution of Color Woodblock Prints in Ukiyoe”
• Adachi Institute of Woodcut Prints, “Tracing the Evolution of Edo Prints Through the Color Red”

The Research Phase That Changed Everything

Before writing any code, I dove into research. This wasn’t planned as a three-week endeavor, but that’s what it became.

I studied over 25 websites: museum archives, printmaking workshops, academic papers, kimono shops that explained traditional dyeing techniques. At some point, looking at my Figma board full of screenshots and reference links, I literally wrote a label that said: “Wow, I looked at a lot of stuff.”

Why the research mattered

The critical discovery was understanding the exact printing process. A single nishiki-e print could require 10 or more separate woodblocks, each carrying a different pigment. The printer would apply them in a specific order, typically starting with the lightest colors and building toward darker ones.

This wasn’t trivia. It directly determined my animation design: which color layer appears first, which comes second, and how they stack. Without understanding the process, I would have animated layers in arbitrary order. Technically functional, but culturally wrong.

Further details are explained on this page:

浮世絵はどうやって作られる？ 「浮世絵きほんのき！」vol.3｜コラム｜きものと（着物メディア）│きものが紡ぐ豊かな物語。－京都きもの市場

Color extraction

Using Python, I extracted dominant color palettes from actual prints across different eras. The data told a clear story: sumizuri-e uses one color (black). Benizuri-e adds two or three. Nishiki-e explodes into a full spectrum.

I visualized this evolution using a normalized 9-slot color system. Each slot represents a color category. In early eras, most slots are “ghosts,” faint placeholders showing colors that didn’t yet exist in the printing technology. As you move through the timeline, slots light up one by one.

Where AI Helped, and Where It Didn’t

What AI accelerated

AI was genuinely powerful for certain tasks:

Code scaffolding. Component structures, responsive layouts, animation parameter tuning. AI handled these efficiently. What might take me an hour of boilerplate, AI generated in minutes.

Exploring variations. When I wanted to compare layout approaches like vertical stacking vs. horizontal scroll vs. isometric view, AI could rapidly generate multiple options for comparison.

Debugging and optimization. CSS specificity conflicts, Framer Motion configurations, scroll behavior issues. AI was a capable pair-programming partner.

Bilingual content. The site supports Japanese and English. AI helped with translation and tone adjustment across both languages.

What AI couldn’t do

Hand-coloring layer assets. This was the biggest surprise. The core visual experience (color layers stacking to form a complete print) required images of each individual color layer. These don’t exist in any dataset. I tried AI image generation, but it couldn’t produce layers with the precise color matching and edge accuracy needed.

I ended up spending weeks in Adobe Fresco, manually coloring each layer by referencing the original prints.

❤️ Red pigment sections on one layer.

💚 Green foliage on another.

💛 Gold accents on a third.

This was the most time-consuming part of the entire project, and AI couldn’t shortcut it.

平塗り sumizuri-BEST-FINAL-reーGIMP-15x.mp4

Note: the base ukiyoe illustrations themselves were AI-generated. What required my hand was the separation of colors into individual layers. The “coloring” work that mimics what a real Edo-period printer would do with separate woodblocks.

Cultural design judgments. Ukiyoe was folk art, mass-produced and casually consumed during the Edo period. It was never displayed in museum frames. This insight shaped my gallery design: more organic, less sterile. AI couldn’t have made this judgment because it requires understanding the cultural context beyond the visual surface.

Creative leaps. The decision to use isometric perspective came from a casual observation I jotted in Figma: “Hmm… Japanese paintings are so isometric. What’s the name?” This wasn’t a logical deduction. It was a human noticing a pattern. AI excels at executing within defined parameters. It doesn’t wander.

The Design Journey

Top page: Arriving at isometric

The layer animation went through multiple phases of iteration:

1. Full-width side-by-side layout

2. Media player-style timeline controls

3. AI-generated pattern exploration and comparison

4. Exploring ways to show the layering process at once

5. Final: isometric view

The isometric choice wasn’t arbitrary. Many ukiyoe prints feature a bird’s-eye, almost planimetric composition, very different from Western perspective.

The Ota Memorial Museum of Art describes how Edo-period artists depicted cityscapes from impossible overhead viewpoints, using ”only imagination and paintbrush to fix their gaze from high above.”

That flat, elevated viewpoint is essentially isometric. It felt right to present the layers the same way.

The need for storytelling

Beautiful animations without context confuse first-time visitors. “Why are colored shapes stacking on top of each other?” After recognizing this problem, I added explanatory text sections for each technique era, using horizontal scroll to avoid excessive vertical length.

Timeline: Learning from architectural failure

The timeline page taught me the hardest lesson of the project.

My initial design captured wheel events and converted vertical scrolling into horizontal movement. It looked impressive in demos but fought against native browser scrolling. Users couldn’t reliably navigate between sections. I patched it twice, adjusting scroll sensitivity and adding cooldown timers. Both patches failed.

On the third attempt, I stopped patching and rebuilt the entire page using native vertical scroll with parallax effects. Completely different architecture.

The lesson was clear: Don’t patch architectural mismatches. If the design approach is fundamentally wrong, redesign from the foundation.

Logo: From hanko stamps to hanmoto marks

The logo exploration was a microcosm of the project’s tension between tradition and modernity.

I started with hanko (seal stamp) aesthetics combined with the characters for “浮世絵.” But seal-framed logos had poor readability. Simplifying too much lost the Edo-period feel.

The breakthrough came from researching hanmoto (版元), Edo-period publishers.

Every print carried their publisher’s mark: a small geometric logo that was essentially a brand identity, centuries before the term existed. These marks used simple geometric motifs like mountain shapes, diamonds, and circles, while carrying rich cultural meaning.

This reference defined the final direction. The logo uses three diamond shapes stacked vertically, representing:

• Black (bottom): the ink of sumizuri-e
• Crimson (middle): the red of benizuri-e
• Indigo (top): the blue of nishiki-e

Running through all three is a vertical line evoking hangi (版木), the woodblock. The foundation from which all ukiyoe prints are born.

I tried several variations, including circles and diamond-circle combinations, before settling on the diamond-plus-woodblock composition as the one that best captured both the strength of hanmoto marks and the spirit of ukiyoe.

Behind the Scenes: Working with AI

Talking to Cursor

Most of my interaction with the AI coding assistant (Cursor) was through voice input rather than typing. Speaking my thoughts let me convey subtle design preferences and nuances that I might skip when typing.

Template prompts

I created four reusable prompt templates for different situations based on this X article:

1. New project initialization
2. Integrating into an existing project
3. Specific bug fixing
4. Pre-launch quality check

The fourth template generated a comprehensive checklist that included Lighthouse testing, which the AI told me to run manually in the browser console. In a way, the AI was managing the human.

Lighthouse optimization

After the AI flagged performance issues, it implemented fixes including GPU-composited shimmer animations, WCAG AA contrast compliance, ARIA role corrections, and LCP image priority adjustments. The site went from failing several audits to passing across the board.

Preparing data in Figma

To write this article and update my portfolio, I organized all the scattered design explorations and reference materials in Figma. I created labeled sections so that when I fed the Figma data to the AI assistant, it could read not just the visual layout but the meaning behind each design decision.

The prompt I gave was deliberately rough: a voice-input dump of everything I wanted done, organized with headers.

# What I Want to Do

I've written down what I want to do below in a rough way. Rather than jumping straight into execution, I'd like you to first understand each item and figure out how to organize them into actual actions. If anything I wrote is unclear, please ask me questions so we can create a higher-resolution plan.

## Add to Portfolio

I want to add this as a project to my portfolio.
The portfolio projects are here: `…/portfolio`

## I gathered things I've done so far that might be useful when creating the portfolio

In the following Figma section, I've collected the main parts of my design process so far (not everything, but the key ones). This is a pretty large dataset, so you don't need to go all the way down to the deepest layers. I've added meaning to the layer names, so start by understanding those.
https://www.figma.com/design/…

## Turn It into Articles

I also want to write articles in Japanese (note) and English (Medium). The writing guidelines for each platform are here: `…note-article/_guidelines`
Also, please look at my existing articles to learn my writing style and characteristics.
https://note.com/io_73
https://medium.com/@iori73

The Cursor then executed four parallel tasks: portfolio page creation, Japanese article draft, English article draft, and image asset organization.

What This Means for Designers Working with AI

This project crystallized something I’d been sensing:

AI can create, but it can’t craft.

AI isn’t limited to well-defined tasks. It can research ambiguous topics, combine ideas in novel ways, and generate designs you didn’t ask for. It’s more capable than most people give it credit for.

But what AI produces tends to lack something I’d call craft: the narrative behind a design choice, the reason a particular color was picked over another, the obsessive attention to a detail that nobody asked for but that makes the whole thing feel intentional.

When I chose to base the logo on Edo-period publisher marks, that decision carried a story. When I used isometric perspective because Japanese painting is inherently flat, that connection came from weeks of looking at prints and noticing a pattern. AI could have generated an isometric layout, but it wouldn’t have known why it was the right choice for this project.

The most important creative decisions came from:

1. Immersive research: spending weeks studying printing techniques before writing code
2. Unexpected connections: linking the flat spatial composition of Japanese painting to isometric animation
3. Cultural sensitivity: understanding that ukiyoe was folk art, not fine art, and letting that shape the design
4. Accepting limitations: recognizing when AI generation wasn’t good enough and choosing to hand-color instead
5. Learning from failure: when two patches didn’t fix the scroll problem, stopping and redesigning the architecture

AI could assist with all of these. But the narrative that ties them together, the sense that every decision was made for a reason, that’s what makes a project feel crafted. And that’s still human.

The Site

The final site features three main experiences:

• Home page: Scroll-driven layer animations showing how each printing technique works
• Dawn page: Deep dive into three technique eras with layered image compositions
• Timeline: Parallax artwork gallery spanning five historical periods with color palette visualization

Built with React 18, Vite 6, and Framer Motion. CSS design system using color tokens extracted from actual historical pigments. Bilingual in Japanese and English.

Ukiyo-e Timeline — Experience the Evolution of Woodblock Printing

浮世絵タイムライン | Ukiyo-e Timeline

Reference Sites

During this project, I visited 27 websites for research, inspiration, and technical reference. Here is the full list:

Ukiyoe History & Techniques

• Yoshio Kato: Evolution of Color Woodblock Prints
• Yoshio Kato: Ukiyoe Reference Archive
• Adachi Institute of Woodcut Prints: Hokusai Past and Present
• Kimono Ichiba: How Are Ukiyoe Made?
• Touken World Ukiyoe: Differences Between Ukiyoe and Nishiki-e
• Daruma3: Ukiyoe Column
• Yubisha: Ukiyoe Column
• Photo-make: Historical Illustration Archives

Museums & Digital Archives

• Ota Memorial Museum of Art
• Sumida Hokusai Museum
• Fukuda Art Museum
• Tsunagi Art Museum
• Tokyo Museum Collection
• ColBase: National Institutes for Cultural Heritage
• National Diet Library Digital Collections
• Cultural Heritage Online
• Kanazawa Bunko: Reproduction Print Workshop
• Edo Tokyo Kirari

Web Design Inspiration

• The Tawaraya
• Sawataya
• Abemamoru Shouten
• Japanese Botanical Laboratory
• Maneru Design Lab
• Apple iPhone Air page

Other

• Jun’ichirō Tanizaki, In Praise of Shadows (book)

If you’re working on a project where cultural context matters, whether it’s traditional art, local craft, or regional design, I’d love to hear about your experience balancing AI tools with domain-specific knowledge.

Tags: Design Process, Japanese Culture, AI, Web Development, Interaction Design

Content Declaration

- Author: Iori Kawano

- AI Assistance: Yes, AI used for editing and structural suggestions (original draft and all creative decisions by author)

- Date: 2026-02-24

This title was chosen rather spontaneously.

Usually, I ask AI to refine my rough drafts. This time, however, I wrote everything without any AI assistance.

What follows is not a grand claim, but a modest inquiry into my own cognitive tendencies.

Trigger

As a recent graduate navigating early professional life, I once wondered whether I might be incapable of executing tasks strictly as instructed.

After a conversation with my project team, I reflected on my own thinking patterns and wrote the following memo:

My thought tendency?

When the future of a project remains ambiguous and I subconsciously perceive room for my own proposal, I devote excessive cognitive resources to preparing that proposal, deprioritizing the immediate task.

Am I overly assumptive?

( ﾟдﾟ) sudden realization

Does this imply that heavily regulated environments, such as large corporations, are unsuitable for me?

If I simplify the logic, that seems to be the surface conclusion.

Conversely, does this mean I am suited for startups?

Yet I am not a generalist, nor do I possess sufficient experience. (ﾟωﾟ) defeated sigh

That memo remained dormant until I encountered the concept of “novelty exploration” in the work of Nobuko Nakano, a Japanese neuroscientist known for translating behavioral research into accessible personality frameworks.

新版　科学がつきとめた「運のいい人」

What Is Novelty Exploration?

In simplified terms:

Humans uniquely seek uncertainty even when their current environment is safe. Climbing dangerous mountains, diving into deep oceans, or exploring outer space exemplifies this drive.

At a mundane level:

- Choosing a new drink flavor at a convenience store reflects higher novelty exploration.

- Always buying the same oolong tea reflects lower novelty exploration.

- Quickly adopting new electronic devices suggests higher novelty seeking.

- Hesitating to upgrade indicates lower novelty seeking.

Importantly, individuals with low novelty seeking tend to adhere strongly to social norms and rules, sometimes prioritizing them even above personal welfare.

My Self-Assessment

I am not inclined to purchase the newest iPhone, experiment with new foods, or chase emerging music genres.

However, if placed in a toxic workplace, I doubt I would passively comply with unjust practices. I would likely gather legal evidence and confront the issue.

So I asked ChatGPT:

“Within Nakano’s framework of novelty exploration, am I high or low?”

After responding to a series of structured prompts, I received the following analysis.

AI’s Interpretation

The conclusion was neither binary nor simplistic.

I am neither low in novelty exploration nor a high stimulation-driven novelty seeker.

I represent a hybrid cognitive type.

Novelty exploration is not merely liking “new things.” It concerns the locus of motivational energy when encountering uncertainty.

1. Conditional Engagement

I do not pursue novelty unless I perceive meaningful benefit.

This corresponds to instrumental exploration: exploration driven by purpose rather than stimulus.

2. Discomfort with Crowds, Comfort with Unknown Meetups

Crowded stations and commercial complexes induce sensory overload.

Attending an unfamiliar but contextually meaningful meetup does not.

This reflects selective novelty tolerance, not generalized novelty seeking.

Meaningless stimuli are rejected. Meaningful ambiguity is accepted.

3. Cognitive Explosion Under Ambiguity with Agency

When ambiguity implies room for intervention, I become intensely engaged.

This reflects:

• Need for agency
• Cognitive ownership

Rigid rule-bound systems are draining.

Structured systems with redesign potential are stimulating.

4. Low Sensory Novelty, High Conceptual Novelty

My daily habits are consistent. I resist action without clear rationale.

This aligns with low sensory novelty seeking and high harm avoidance.

However, I display strong epistemic exploration: the pursuit of new conceptual connections and abstract meaning structures.

In short:

Low sensory novelty, high conceptual novelty.

Misinterpretation of the Self

Popular images of high novelty seekers emphasize:

• Spontaneity
• Risk-taking
• Rapid behavioral change

My novelty orientation operates internally, within conceptual architecture.

Thus I mistakenly categorized myself as low in novelty exploration.

In reality, my exploration targets meaning rather than environment.

Reinterpreting My Memo

“Am I unsuited for large corporations?”

→ Not necessarily. I am unsuited for environments devoid of intervention space.

“Am I suited for startups?”

→ Early-stage chaos is draining. Mid-stage redesign environments are optimal.

This is not about novelty intensity.

It is about novelty direction.

Proper contextual alignment transforms this trait from liability to strength.

Afterword

The AI’s response did not reveal something radically new.

Rather, it articulated what I had already sensed in tacit form.

It also reminded me that AI does not generate truth.

It generates probabilistic responses based on patterns of expectation.

That distinction remains essential.

Author: Iori

AI Assistance: None in drafting or editing

Date: 2026–02–11

*Disclaimer: The organizations featured in this article were selected based on my personal interest and perspective. This is not intended to be a comprehensive or definitive list.

I’ve compiled 40 Japanese companies and projects working on regional revitalization and environmental issues, organized by category.

This article is for:

• Those interested in rural migration and regional revitalization
• Those seeking careers in environmental sustainability
• Anyone wanting a comprehensive overview of this field

Background

I work as a UI designer. Though I’m not an expert in this field, I’m drawn to keywords like “nature” and “regional development.” I’ve been bookmarking interesting websites in this space.

When I organized them, I had 40! I thought I’d share them.

Category Overview

Category Number Main Areas Regional Revitalization 12 Accommodation, vacant house revitalization, regional economy Natural Environment 18 Biodiversity, forestry, water resources Agriculture 5 Agritech, aquaculture, biotech Architecture & Urban Planning 5 Traditional house renovation, timber utilization

Organizing 40 Sites on Two Axes

Looking at these 40 sites, you might wonder, “Where should I start?”

So I organized them along two axes:

• X-axis (Level of Engagement): From “View & Learn” to “Visit & Experience”
• Y-axis (Approach): From “Technology & Digital” to “Community-focused”

If you want to start with information gathering, look at the left side. If you want to visit places, look at the right. If you’re interested in digital tools, check the bottom. If you’re interested in direct community engagement, look at the top.

How to Read the Graph:

• X-axis (Engagement): Left to right “View & Learn” → “Buy & Use” → “Visit & Experience”
• Y-axis (Approach): Bottom to top “Technology & Digital” → “Business & Enterprise” → “Community-focused”

Four Quadrants:

• Bottom-left: Technology × Information (RAIDA, BIOME, Takram Cast, etc.)
• Bottom-right: Technology × Experience (WOTA, BLUE OCEAN DOME, etc.)
• Top-left: Community × Information (minto., Kikoeru Iwate, etc.)
• Top-right: Community × Experience (NIPPONIA Kosuge, Satoyama Jujo, KURKKU FIELDS, etc.)

Regional Revitalization (12 Organizations)

Organizations rediscovering rural charm, revitalizing vacant houses, and visualizing regional economies.

Regional Development × Accommodation

• NIPPONIA Kosuge Genryu no Mura — A distributed hotel renovating traditional houses in Kosuge Village, Yamanashi. The concept of “the entire village as a hotel” contributes to regional revitalization. https://nipponia-kosuge.jp/

• SUIDEN TERRASSE — A hotel “floating” on rice fields in Tsuruoka, Yamagata. Designed by architect Shigeru Ban, it offers accommodation experiences that leverage the rice field landscape. https://suiden-terrasse.com/

• Satoyama Jujo — A renovated traditional inn in Minamiuonuma, Niigata. Designer furniture and art create a space offering local gastronomy. https://www.satoyama-jujo.com/

Regional Development & Soft Developer

• Staple Inc. — A “soft developer based in rural and urban areas.” An approach to developing regions through software (content and experience) rather than hardware. https://staplejp.com/

• mount inc. — Handles numerous projects to draw out regional appeal. https://mount.jp/

Regional Economic Data & Analysis

• RAIDA (Regional Revitalization Data Analysis Platform) — A data analysis platform provided by the Cabinet Office Regional Revitalization Office. Supports understanding regional issues through data. Successor to V-RESAS. https://raida.go.jp/

• LEDIX (Teikoku Databank) — A website visualizing regional economic stakeholders through data. Created by Takram with Teikoku Databank in 2017. https://www.takram.com/ja/projects/ledix

Regional Food & Industrial DX

• Uuuo Inc. — DX for the fishing industry. A platform connecting fishermen and restaurants, aiming to streamline distribution and improve fishermen’s revenue. https://uuuo.co.jp/

Vacant House Revitalization & Migration Support

• minto. | Hiroshima Vacant House Bank — Hiroshima Prefecture’s vacant house information platform. Clean UI design, easy to use for migration consideration. https://minto-hiroshima.jp/

Regional Culture & Art

• Tsunagi Art Museum — A municipal art museum in Tsunagi Town, Kumamoto. Has been implementing resident-participatory art projects since 1984 with the concept of “viewing the entire town as a museum.” https://www.tsunagi-art.jp/

• Edo Tokyo Kirari Project — A traditional craft support project promoted by Tokyo Metropolitan Government. https://edotokyokirari.jp/

• Kikoeru Iwate — Iwate Town’s SDGs Future City Project. https://iwatetown-sdgs.jp/

Natural Environment (18 Organizations)

Companies and projects rethinking the relationship between nature and humans, and addressing environmental issues. 18 of the 40 are concentrated in this category.

Biodiversity × Technology

• BIOME Inc. — A biodiversity platform. Provides an app that identifies creatures just by taking smartphone photos. Realizes citizen-participatory ecosystem research through gamification. https://biome.co.jp/

• mui Lab — Develops “mui,” a product where a wooden board becomes an interface. Design philosophy of embedding technology into natural materials. https://muilab.com/

• Convivial Technology — Brings Ivan Illich’s philosophy to the present. Explores “a society where humans and technology coexist.” https://convivial.tech/

Water Resources & Water Cycle

• WOTA Corporation — Developed technology that can reuse 98% of water. With the vision “Water Freedom for Everyone, Everywhere,” promotes use in disasters and areas with limited water resources. https://wota.co.jp/

Outdoor & Field Research

• Goldwin Inc. — An outdoor company operating THE NORTH FACE and other brands. Places sustainability initiatives at the center of business. https://about.goldwin.co.jp/

• Goldwin Field Research Lab. — Goldwin’s research institution. Researches and disseminates changes occurring in the field. https://fieldresearchlab.goldwin.co.jp/

Forest Resources & Forestry

• Hidakuma (Hida no Mori de Kuma wa Odoru) — A project utilizing broadleaf trees from Hida. Collaborates with architects and designers on initiatives connecting forests and cities. https://hidakuma.com/

• JINEN Corporation — Concept is “living in nature.” https://jinen.earth/

• Japan Plant Research Institute — Discovers the potential of plants from mountains across Japan as food ingredients. Conducts research and product development to utilize Japanese mountain materials in food. https://nihonkusakilab.com/

• Aobagumi Corporation — A forestry and reforestation company. With the vision of “growing natural capital,” engages in forest creation and forest regeneration, including ski slope restoration projects. https://aobagumi.jp/

• Miyazawa Timber Industry Co., Ltd. — Handles everything from forest maintenance to wood biomass power generation in Nagano Prefecture. Expanding the possibilities of timber. https://www.mwwi.co.jp/

• Healthian Wood — A project in Tateyama Town, Toyama Prefecture. Operates facilities connecting forests and health. https://healthian-wood.jp/

Environment & Sustainability

• ThreeTransitions.earth — Proposes three transitions (energy, food, nature) for the restoration of nature and humanity. Design firm Takram is involved. https://www.threetransitions.earth/

• NOSIGNER — A design firm with “Evolutionary Design” as its motto. Addresses environmental issues through design approaches learning from biological evolution. https://nosigner.com/

• BLUE OCEAN DOME — A pavilion at the 2025 Osaka Expo. Plans exhibitions on the future of marine environments. https://zeri.jp/expo2025/

• Aso Grassland Restoration Project — An initiative to protect Aso’s grasslands. Inherits traditional management methods such as controlled burning. https://www.asogreenstock.com/sougensaisei/

Wildlife & Nature Experience

• NPO Picchio — Researches and practices coexistence with wildlife in Karuizawa. Known particularly for Asian black bear management. https://npo.picchio.jp/

Podcast & Media

• Takram Cast — Design firm Takram’s podcast. Also develops discussions on transitions between nature and humanity. https://www.takram.com/ja/podcasts

Agriculture & Food (5 Organizations)

Companies thinking about the future of food and agriculture. Notable for combining technology with agriculture and fisheries.

• Chitose Group — A group of companies addressing environmental issues through biotechnology. Conducts algae research. https://chitose-bio.com/jp/

• Nihon Agri Inc. — Transforms agriculture into a growth industry by exporting Japanese agricultural products globally. Handles everything from large-scale orchard management to packing facility automation/DX and global expansion. https://nihon-agri.com/

• UMITRON — “Implementing sustainable aquaculture on Earth.” Optimizes fish farming with AI, reducing feed waste. https://umitron.com/ja/index.html

• IntegriCulture Inc. — Creates the future of food through cell culture technology. Develops cultured meat. https://integriculture.com/

• KURKKU FIELDS — A complex facility for “people, agriculture, food, and art” in Kisarazu, Chiba. Led by music producer Takeshi Kobayashi. https://kurkkufields.jp/

Architecture & Urban Planning (5 Organizations)

Companies attempting to reconstruct relationships with regions and nature through architecture and design.

• VUILD — Democratizing architecture through digital fabrication. Building a system where anyone can build houses using local timber. https://vuild.co.jp/

• Sansuisha — Store design utilizing old timber from traditional houses. Based in Nagano Prefecture, active nationwide. https://www.sansui-sha.jp/

• Namba Kenchiku Kobo — A construction company handling traditional house renovation in Okayama Prefecture. https://www.nanbakenchiku.co.jp/

• Takenaka Carpentry Tools Museum | Plants × Craftsmanship — An exhibition project considering the relationship between plants and craftsmanship. https://plantsandcrafts.dougukan.jp/

• Design Studio-L — A branding design company in Nagano Prefecture. Handles branding for numerous regional companies. https://designstudio-l.jp/

Three Trends from Organizing 40 Sites

1. “Nature with Humans” Not “Nature vs. Humans”

I felt many companies are seeking ways for nature and humans to coexist, rather than the confrontational structure of “environmental protection = restricting human activities.”

2. Technology as a Means, Not an End

While many companies utilize AI and IoT, technology itself is not the goal. It’s positioned as a means to solve challenges related to nature and regions.

3. The Power of Design to Convey Vision

As a UI designer, I noticed that companies with beautiful websites have a strong will to convey what they do. Design functions as a means to communicate vision.

Future Plans

• Want to actually visit places like NIPPONIA Kosuge and KURKKU FIELDS
• Want to research these companies’ business models and write another article
• Want to incorporate these perspectives into my own design work

Thank you for reading this long article to the end!

If you know of other interesting sites, please share in the comments!

I’d love to connect with people working in, active in, or interested in the fields introduced in this article. Please feel free to reach out through likes and comments! 🙇‍♀️

My understanding is extremely low, but I like the concept of emptiness proposed by Kenya Hara, which is connected to the Buddhist idea of 「空 kū」 (emptiness). I was thinking about what other examples of such emptiness might exist.

In this article, rather than prioritizing careful explanation or readability, I want to prioritize a record of my thoughts in their raw state, along with my clumsy vocabulary.

Pine Trees Screen

Pine Trees Screen (Shōrin-zu byōbu).

A Japanese favorite: Pine Trees Screen.

There are many explanations on the internet, so here I will quote Google’s AI Overview.

Google search: “Pine Trees Screen space”

The spatial expression in Hasegawa Tōhaku’s Pine Trees Screen is characterized by the skillful use of blank space (emptiness), creating an infinite sense of expanse and depth through mist, haze, air, and earth.

Unlike conventional landscape paintings that clearly distinguish foreground, middle ground, and background, Tōhaku makes pine trees “emerge” through variations in ink density and empty space.

This innovative expression produces an abstract, mysterious, and infinite space that stimulates the viewer’s imagination.

Things Inside a Shrine

Inside four stakes placed at the corners.

Thin stakes are placed at the four corners, and a rope is stretched between them. By doing so, an empty unit, that is, an empty unit, is created.

Emptiness exists as the very possibility of being filled. A god might notice it and wander in. Gods are always quick to notice such things. They will not necessarily enter, but they might. That possibility, that “might,” is what ancient Japanese people put their hands together for.

https://www.muji.net/lab/report/100203-01.html

This may be slightly off-topic, but it reminded me of the phenomenon where, if you prepare a frame, a cat somehow enters it. The tendency to direct attention toward empty space might be something animalistic.

Behind the Sacred Door

This is a personal experience, but when I was in elementary school, I saw the door of a nearby shrine open for the first time.

At the time, I believed that the door was not meant to be opened, and that only special people were allowed to see what was inside. So when it opened so easily, I did not feel disappointed so much as I felt my strength drain away.

There was nothing inside that could be called the main object. The space itself occupied most of what lay beyond the door. Still, in that dim space surrounded by light-colored wood, I felt that there might be something different from the everyday world.
(It is also possible that there was nothing inside simply because it happened to be a moment of replacement or cleaning…)

Between Houses

Even a gap too small for a person to enter feels like it contains something.

Even spaces that people cannot enter feel as though something might exist there.

In particular, I really liked the third photograph in the section “Spaces also used as storage” from the wonderful blog Gaps Between Buildings.

I found it while searching for reference images for this article. Since reproduction without permission is prohibited, I cannot include it here.

We rarely view such spaces as public objects, but for the remaining items, they are perfectly adequate places to live. Plants grow thickly there as well, and it makes me want to cheer them on (?).

The following description was also wonderful, and it made me feel a lively, distinctly Japanese sense of clutter.

Flowerpots, blocks, and pieces of wood are placed in a disorderly manner, and many weeds grow there. Because such gaps are rarely seen by people, they are often poorly managed.

Still, I find this scenery extremely appealing, as it seems to express something very Japanese.

(№327, Tokyo)

From this photograph, it does not look as though disposable items were simply thrown away. The pots and concrete slabs appear to have been placed along the steps.

If I were to arbitrarily guess the mindset of the person who placed them there, it might have gone something like this:
“I don’t need this, but throwing it away is a hassle. Oh, there’s a convenient gap here. Still, it feels wrong to just toss it carelessly, so I’ll line it up along the steps.”

Regarding this tendency of Japanese people to create environments filled with various objects, I heard an interesting phrase in the following podcast, so I want to share it here.

What was discussed was that the “cluttered” feeling of Japanese pop culture and websites is not simply disorder, but intentionally organized chaos. That made sense to me.

Under the Eaves

A mysterious space that everyone loves (?) and that also appears in My Neighbor Totoro.

To create this atmosphere that feels like something is there, naturally formed shadows are also necessary. There’s a quiet dignity to it, though I’m not sure how to put it into words.

These were my scattered mental pieces, written down freely and enjoyably, just as they came to me.

If you have your own thoughts about emptiness or leftover space, I would be happy to read them in the comments.

Thanks for reading!

Recently, I was listening to a G1 by GLOBIS podcast titled “The Heart of Regional Revitalization: ‘Deep Local is Global’ — How Unique Local Charm Attracts the World!” when I came across this interesting statement:

“Most hit manga artists who have debuted are from rural areas. The perspectives and ways of thinking unique to people from the countryside are deeply embedded in their manga. Gratitude toward others and compassion for people are really important in manga.”

— Shuichi Mochida, Coamix Inc.

I found this fascinating, but I also listened with some skepticism.

After all, there are plenty of creators from Tokyo too, and the discussion seemed to intentionally gloss over that fact (understandably, given the “regional revitalization” context). I have no intention of dismissing the panelists, but I didn’t want to take their claims at face value.

By the way, Mochida’s comments were specifically about manga artists. The podcast did mention Studio Ghibli, but the argument about rural origins being advantageous was primarily made in the context of manga artists.

So I thought: What would I see if I actually visualized this data?

I decided to include both manga artists and anime directors in my visualization. I built it in about an hour using Cursor, and here’s how it went.

The Final Product

Here’s the completed project:

🔗 Site URL: https://manga-anime-authors.vercel.app/

Note: The visual design, such as typography, colors, spacing, and layout, is entirely AI-generated output. I intentionally made no modifications to preserve a record of what Cursor produced on its own.

I plotted the birthplaces of 40 representative Japanese manga artists and anime directors on a map of Japan.

(Why 40? → When Cursor asked “What kind of creators do you want to include?”, I answered “famous ones,” and the AI automatically selected 40 people.)

Manga artists: Osamu Tezuka, Akira Toriyama, Eiichiro Oda, Takehiko Inoue…

Anime directors: Hayao Miyazaki, Hideaki Anno, Makoto Shinkai, Mamoru Hosoda…

Click on any creator to see detailed information (birth year, active decades, major works).

The Motivation

“Are Japanese creators actually more likely to come from outside Tokyo?”

To verify this, collecting and visualizing data seemed like the fastest approach. So I decided to build it with Cursor (Plan mode, Opus 4.5).

The Initial Prompt

Rather than jumping straight into code, I first had Cursor help me organize the requirements. I dictated this prompt using voice input:

I heard that many Japanese manga artists come from rural areas rather than
cities like Tokyo. I've heard that what makes Japanese content interesting
compared to foreign content is that creators grew up with rich lives in
the countryside and developed rich spirits.

I want to visualize the birthplaces of Japanese manga artists and anime
creators on a map of Japan.
First, let's start by listing representative Japanese manga artists and
anime creators, then think about what technology to use and what kind of
visual to create.

Also, I think the active periods and birth years differ among creators,
so I want to consider that too. It would be nice to have filtering
functionality for that.

After sending this prompt, Cursor worked out everything from technology selection to proposing a list of creators.

Technology Stack

The stack Cursor proposed:

• React — UI framework
• D3.js — Map rendering & interactions
• Vite — Build tool
• Vercel — Hosting

D3.js is commonly used for map visualization. It draws the Japan map using SVG and places markers for each prefecture.

What I Did During Development

1. Data Collection

I organized data for 40 creators in JSON format:

{
  "id": 1,
  "name": "手塚治虫",
  "nameRomaji": "Tezuka Osamu",
  "birthYear": 1928,
  "prefecture": "大阪府",
  "category": "漫画家",
  "works": [
    {"title": "鉄腕アトム", "year": 1952},
    {"title": "ブラック・ジャック", "year": 1973}
  ]
}

2. Fetching Map Data

I set it up to fetch map data from external sources. Looking at the code later, I noticed that Cursor had automatically implemented “fallback to multiple sources”:

const sources = [
  { name: 'world-atlas countries (TopoJSON)', url: '...' },
  { name: 'world-atlas countries detailed', url: '...' },
  { name: 'dataofjapan land (GeoJSON)', url: '...' }
];

// Try multiple sources in order, use whichever succeeds
for (const source of sources) {
  try {
    const response = await fetch(source.url);
    // ...
  } catch (error) {
    console.warn(`Failed to load from ${source.name}:`, error);
  }
}

Apparently it’s designed to automatically fall back to another source if one fails.

3. Overlapping Marker Problem

When multiple creators are from the same prefecture, their markers overlap.

For example, just from Tokyo:

• Hayao Miyazaki
• Yoichi Takahashi
• Masami Kurumada
• Naoki Urasawa
• Mamoru Oshii
• Tetsuo Hara
• Chika Umino

In the first output, Tokyo-based creators’ markers overlapped, making some impossible to select.

So I used D3.js force simulation to slightly offset markers so they don’t overlap:

const simulation = d3.forceSimulation(nodes)
  .force('collision', d3.forceCollide()
    .radius(d => d.radius)
    .strength(0.3)
  )
  .alphaDecay(0.3); // Converge quickly

Now multiple markers from the same prefecture are neatly arranged.

4. Zoom UX Problem

I added zoom functionality to the map, but initially page scrolling and map zooming got mixed up.

When users tried to scroll the page, the map would zoom instead… terrible UX.

Following the pattern of embedded Google Maps, I modified it so mouse wheel only zooms with Ctrl/Cmd + scroll. On MacBook trackpads, two-finger pinch zoom works directly (because the browser internally treats it as Ctrl key).

const zoom = d3.zoom()
  .filter((event) => {
    if (event.type === 'wheel') {
      // Only zoom when Ctrl/Cmd is pressed
      if (event.ctrlKey || event.metaKey) {
        event.preventDefault();
        return true;
      }
      return false; // Normal scroll goes to page scroll
    }
    // ...
  });

Patterns That Emerged

Visualizing the data revealed some interesting trends:

More Rural Origins Than I Expected

Kyushu: Eiichiro Oda (Kumamoto), Takehiko Inoue (Kagoshima), Koyoharu Gotouge (Fukuoka), Moto Hagio (Fukuoka)…

Tohoku: Yoshihiro Togashi (Yamagata), Hirohiko Araki (Miyagi), Tatsuki Fujimoto (Akita), Gege Akutami (Iwate)…

Hokkaido: Hideaki Sorachi, Hiromu Arakawa, Satoshi Kon

However, looking at the actual rankings, Tokyo was #1. There are indeed many Tokyo-born creators.

But from 2nd place onward, they’re scattered across the country. Many creators come from “rural” areas like Kyushu, Tohoku, and Hokkaido.

It’s not as simple as “rural origins produce interesting works” — talent emerges from both Tokyo and rural areas. That’s my honest takeaway from the data.

One part of the podcast that stuck with me:

Katsuhiro Otomo (AKIRA) was from deep rural Miyagi with such strong longing for the city that he drew those skyscraper skylines.

Hajime Isayama (Attack on Titan) was from Oita, surrounded by mountains, thinking “what’s beyond those mountains?” — and those mountains became the walls.

Eiichiro Oda (ONE PIECE) wouldn’t have created that work without Kumamoto’s “wasamon culture” (a local spirit of loving new things and wanting to be the best in the world).

The perspective that environments and landscapes from growing up in rural areas directly reflect in their work’s worldview was fascinating.

Thoughts on Using Cursor

What I felt during this development:

What Worked Well

• I could discuss from the requirements stage, thinking together about “what to build”
• Even with complex libraries like D3.js, I could progress while getting sample code
• Rich variety of modes and models to choose from depending on instruction volume and content

What I Kept in Mind

• Start with “let’s think together about what to build” rather than jumping to “build this”
• For UX and animations, AI tends to generate outputs that feel “AI-ish,” so I explained specifically what to improve and how
• Switch to different chats for different topics; don’t stay too long in one chat

Summary

To test the hypothesis that “Japanese manga artists and anime creators are mostly from rural areas,” I built an interactive visualization site with Cursor in about 1 hour.

It’s fascinating how even without deep technical knowledge, you can create something by having a conversation with AI about “I want to build something like this.”

What I’d like to do next:

• Add more creators — Currently only 40, I want to be more comprehensive
• Improve the design — It’s currently AI’s first-pass output; I want visuals that better reflect manga/anime culture
• Summarize insights — Add a section that properly articulates trends and discoveries from the map

I’ll update it when I feel like it.

If you have suggestions like “add this creator” or “I’d like to see this angle,” please let me know in the comments! Thanks for reading!

🔗 Final Site: https://manga-anime-authors.vercel.app/

#AI #Cursor #D3js #React #DataVisualization #Manga #Anime #Japan

Section 1: Introduction

This September, I joined the Shimoda Design Workshop — a five-day program held in a small coastal town on the Izu Peninsula, about three to four hours south of Tokyo.

Organized by Design for Japan, the workshop brings together designers, students, and professionals from around the world to collaborate with the local community on real challenges.

The program was carefully structured: we started with a guided tour of Shimoda and a history lecture at Ryosenji Temple, moved into UX research training and fieldwork with local residents and business owners, then explored Japanese design traditions and mascot culture.

Later in the week, we refined our design proposals while also engaging in discussions with experts, including Zelia Sakhi from Google Japan. Finally, we presented our work to the community and celebrated with local people such as Shimoda city staff and the locals who run international guesthouses.

Participants were divided into four groups, each given a theme with the goal of creating and presenting website design ideas. Team A focused on Activities & Shops, Team B on Restaurants & Food, Team C (my group) on Community & Events, and Team D on Accommodation.

I first learned about the workshop through Tokyo UX Design Meetup, where I had met the organizer. When he invited me, I saw it as my first real chance to design in an international, English-speaking environment while collaborating with people from very different backgrounds. That challenge itself became my main motivation for joining.

For those curious, you can see day-by-day highlights on the official Instagram account.

After the event, the organizer also shared reflections on LinkedIn, giving another perspective on what we achieved together.

From the beginning, I knew this experience would push me out of my comfort zone. And it did — whether in the fast-paced debates with international peers or in unexpected conversations with locals, the workshop expanded my perspective on design and gave me a clearer sense of how I want to grow as a designer.

Section 2: Lessons from Team Discussions

Our team discussions were unlike anything I had experienced before — it was also my very first time discussing design entirely in English. The pace was rapid, almost like machine-gun fire. When one member proposed an idea — “I think this works because of X” — another would immediately counter with “But what if we see it this way?” or “In that case, maybe we could try Y.” Several members of our group in particular drove the momentum forward, and I often found myself fully focused just on following the flow of English.

Even so, there were moments when the group turned to me with “What do you think?” In those instances, I shared what I could — sometimes a cultural context from Japan, other times a small reflection from my perspective.

Compared to others, my voice was less frequent, but I still felt that being Japanese gave me a unique lens.

For example, when the conversation touched on community revitalization in Shimoda, I realized how difficult it must be for my international peers to design for a rural town in a foreign country. I imagined myself in their place — perhaps in an Australian countryside town, since Australia is the country I’m currently most interested in, asked to design tourism strategies — and I deeply respected how they engaged with local culture despite not having that background knowledge.

There were also moments when someone’s idea resonated with me and sparked connections. When one member suggested, “We can highlight the people who love talking with tourists on the web,” it reminded me of Japanese food packaging that often shows the farmer’s face.

I proposed adapting that concept into a website section featuring local business owners, with their portraits, stories, and even their language accessibility levels. The team welcomed it.

Similarly, when another member brought up creating a dashboard for authentic, small events, I suggested embedding Google Calendar so business owners could submit events via a simple Google Form.

These were modest contributions, but they reminded me that my strength might lie in connecting others’ sparks to tangible, culturally grounded solutions.

Another point that struck me was how diverse everyone’s career paths were. Some were between jobs, others were on a gap year, while others were waiting for visas or freelancing. It contrasted strongly with the standardized “new graduate” route in Japan.

Yet despite these different backgrounds, one topic surfaced again and again: uncertainty about AI. It wasn’t necessarily original thinking — we were all reflecting ideas we had absorbed from media or articles — but it showed me how shared anxieties can unify a group.

In the end, I realized that I wasn’t the most active speaker in heated debates, but I learned a lot simply by observing how others exchanged ideas so fluidly. And even in my quieter role, I found small ways to add value.

Section 3: Encounters in Shimoda

Walking through the streets of Shimoda, I was struck by how many foreigners — especially Western-looking tourists and residents — I saw. It felt very different from my daily life in Kawasaki, a city located between Tokyo and Yokohama that is better known for its role as a commuter hub and residential area than as a tourist destination, and gave me a sense of how international the town had become.

One short but memorable encounter was with an elderly man in his 80s. In a brief conversation, he shared perspectives that reflected both the present and the past: the noticeable increase of foreigners in the town, the idea that Japan’s declining population will eventually make immigration unavoidable, and memories rooted in the wartime generation.

Even in a few minutes of exchange, I realized that he did not actively approach or recruit foreigners, yet he did not reject them either — his neutral stance toward foreign tourists surprised me. I had assumed that not only in Shimoda, but generally local residents might not hold positive impressions, so this challenged my own stereotypes.

Moreover, it was the first time I directly heard someone speak about wartime memories from that generation. He himself was born in the year the war began and did not experience it directly, but his father served in the Japanese military. He told me that he first saw his father’s face only when he was in the coffin. That story left a strong impression on me, making me realize how real and tragic the impact of war can be, and how valuable it was to speak with someone who carried such memories within his family.

Locals and city staff told me that in the past few years they have been especially impressed by how polite and respectful foreign tourists have been. They also emphasized that this did not mean foreign tourists had been disrespectful in the past, but rather that recent years left an especially positive impression. I thought this was quite different from the narratives often emphasized in national media.

At the same time, I learned that some foreign residents form circles that remain separate from local society, which reminded me of the complexities behind concepts like “internationalization” or “community building.”

These encounters showed me that design work in such contexts is not just about proposing new ideas, but about listening carefully to the diverse and sometimes conflicting perspectives within a community.

Section 4: Returning to hometown (Personal Note)

Coming back to my hometown, Kawasaki after the workshop, I immediately felt the contrast. Shimoda had been a week of mountains, ocean, and quiet streets, while Kawasaki — a city located between Tokyo and Yokohama that is better known for its role as a commuter hub and residential area than as a tourist destination — offered the dense structures of the city: tall apartments, square buildings, and endless paved surfaces. Even the greenery in the parks felt more like carefully arranged symbols of nature than living ecosystems.

Lately I have been immersing myself in the books and lectures of anatomist and essayist Takeshi Yoro, and his thinking has deeply influenced my mind. He often says that cities are products of consciousness — that urban spaces are essentially extensions of the brain. Reflecting on his words, I felt how narrow Japanese cities in particular can appear abnormal for human life, and this struck me strongly upon returning from Shimoda.

This reflection wasn’t part of the workshop’s official theme — it was more personal. Yet, it helped me recognize what kinds of environments energize me as a designer. I realized that I am drawn to projects where design is not isolated from nature, but instead connected to real communities, physical places, and the everyday lives of people. I also felt uneasy about designing products whose primary goal is to “hack” user psychology to drive spending or maximize time-on-site. I don’t want my work to prioritize extracting attention or revenue at the cost of people’s well‑being; I prefer design that respects and supports communities rather than manipulates them.

Section 5: Conclusion & Looking Forward

The Shimoda Design Workshop was more than just a week of design exercises — it was a turning point in how I think about collaboration, culture, and my own role as a designer. Working in a multicultural team showed me how important it is to share context clearly and to build confidence in participating actively, even when language feels like a barrier.

At the same time, I discovered that I may have a strength in connecting sparks of ideas from others and translating them into tangible, culturally grounded proposals.

Looking ahead, I want to continue developing two key abilities: the skill of collaborating across cultures and the ability to shape scattered ideas into concrete outcomes. These are the foundations I want to carry into my career, whether working with international teams, local communities, or future projects that bridge technology and society.

Most importantly, Shimoda reminded me that design is not just about interfaces or outputs — it is about people, histories, and the values that guide our choices. I want to keep building a career that respects those dimensions, and to take on projects that bring design closer to real communities rather than further away from them.