2025,江端さんの忘備録

BS世界のドキュメンタリー シリーズ ウォルト・ディズニー (4)「“夢の国”ディズニーランド」を見ました。

I watched the BS World Documentary Series Walt Disney (4)  ‘Land of Dreams’ Disneyland.

この番組の中で、ウォルト・ディズニーが『我々が長年築き上げたものであり、何かを象徴するものだ』という発言をしていたという場面が出てきます。

In this program, there is a scene where Walt Disney is quoted as saying, “It's something we've built up over the years, and it represents something.”

これは、"ディズニー"という言葉が、『1日楽しめば次の日に忘れるものではなく、世代を超えて伝わっていく概念となるもの』と解説されていました。

The word “Disney” was explained as “something that is not something that you forget the next day after having fun for a day, but is a concept passed down from generation to generation.”

で、彼は、これを「シンボル(象徴)」と言ったそうです。当時は「ブランド」という言葉がなかったからだそうです。

He called it a “symbol” because the word “brand” did not exist then.

-----

現在、日本人の少なくない人が、現在の日本国憲法を「米国から押し付けられた憲法である」と批判しております。

Many Japanese people criticize the current Japanese Constitution imposed on Japan by the United States.”

特に、日本国憲法第1条 の「天皇は日本の象徴」については、『法律に用いるフレーズではない』とまで言い切る人もいた、という話も聞いたことがあります(但し、出所は辿れませんでしたので、単なる風評かもしれません)。

In particular, I have heard that some people even go as far as to say that the phrase “The Emperor is the symbol of Japan” in Article 1 of the Japanese Constitution “is not a phrase used in-laws” (however, I was unable to trace the source, so it may just be a rumor).

で、今回の番組を見ているうちに、「ブランド」という言葉のない時代において、「シンボル=象徴」が『我々が長年築き上げたものであり、何かを象徴するもの』と考えれば、日本国憲法第1条の「象徴天皇」は、(法律として成文しているかどうかはさておき)見事な表現である、とも言えるのではないか、と思いました。

As I watched this program, I thought that if we think of a symbol as something that represents something we have built up over many years, then the “symbolic emperor” in Article 1 of the Japanese Constitution (regardless of whether or not it is written as a law) is a brilliant expression.

これは、米国から押しつけられた憲法の第1条が、『日本という国家は、我々日本人が長年築き上げたものであり、天皇とはそれを象徴するものだ』という解釈が成立し得るからです。

This is because the first article of the Constitution, which the United States imposed on Japan, can be interpreted as meaning that “the nation of Japan was built up over many years by the Japanese people, and the Emperor symbolizes this.”

米国が内戦で分裂して四散して飛散しても、日本国は未来永劫に続く、ということを、他国(米国)が担保してくれた ―― という、超善意的かつ、超楽天的解釈ができるのではないでしょうか ―― ちがうか?

Even if the United States is divided and scattered by civil war, the other country (the United States) has guaranteed that the Japanese nation will continue forever. Isn't that a super-optimistic and super-benign interpretation?

-----

で、ちょっとここから戯言なんですが。

Now, I'm going to start talking nonsense.

当時、「ブランド」という言葉がなくて、やむなく「シンボル(象徴)」を使ったとするのであれば、本来の日本国憲法第1条は、

If we assume that the word “brand” did not exist at the time and that “symbol” was used as a substitute, then Article 1 of the original Japanese Constitution could be read as

―― 天皇は日本のブランドである(修正憲法第1条)

“The Emperor is the brand of Japan(Article 1 of the revised Constitution)

と読み替え可能か、検討してみました。

いかんな。このままでは、天皇が日本国が所有する物品(バックとか、自動車とか)のように読める。

This isn't good. If it stays like this, it will read like the Emperor owns the Japanese nation (like a bag or a car).

では、

Then, 

―― 天皇は日本人を表現するブランドである

the Emperor is a brand that expresses the Japanese people.

うん、これもちょっとしっくりこんなあ。

Yeah, this doesn't fit a little.

なら、

Next, 

―― 天皇は日本人の誇りを象徴するブランドである

the Emperor is a brand that symbolizes the pride of the Japanese people.

これは賛否が分かりそうです。天皇制を支持しない人もいるからです(私は、天皇制は「日本国最高位の与信システム」と認識しています)。

It seems that there will be both supporters and opponents of this idea because some people do not support the Emperor System (I recognize it as “the highest credit system in Japan”).

(↑コラムに飛びます)

(Jump to the column)

----

「ブランド」という言葉には、商業的な価値やマーケティング要素が色濃く含まれるため、やはり天皇の地位を表現するには違和感が残りますね。

The word “brand” has a substantial commercial value and marketing element, so using it to describe the Emperor's position still feels strange.

「象徴」という言葉が持つ、政治的・文化的な中立性や歴史的背景を考えると、やはり日本国憲法第1条の表現は、絶妙なバランスのもとに成り立っているように思えます。

Considering the political and cultural neutrality and historical background of the word “symbol,” it seems that the wording of Article 1 of the Japanese Constitution is based on a delicate balance.

つまり、「象徴」という表現は、単なる代替語の問題ではなく、日本の歴史や文化の中での天皇の役割を最も適切に表現するために選ばれた言葉であると言えると思います。

In other words, I think that the expression “symbol” is not just a matter of choosing a substitute word but a word that was chosen to most appropriately express the role of the Emperor in Japanese history and culture.

もし「ブランド」という言葉が当時存在していたとしても、憲法に採用される可能性は低かったかもしれんなぁ、と考えています。

Even if the word “brand” had existed then, I think it would have been unlikely to be adopted into the constitution.

私は、日本の神様(天皇陛下のご先祖)が、結構「ポンコツ」だったという話が、結構好きです。

2025,江端さんの技術メモ

TC帯域を設定して、レイテンシーを設定した後、パケロスや遅延を設定する場合、だいたい、こうやるのですが、

sudo tc qdisc add dev eth0 root handle 1: tbf rate 10mbit burst 10kb latency 500ms
sudo tc qdisc add dev eth0 parent 1:1 handle 10: netem delay 10000ms loss 1%

RTNETLINK answers: No such file or directory と言われます。

Webで調べてみたところ、山のように、同じ"悲鳴"が飛んでいました。

で、解決方ですが、ChatGPTが、逆転の発想のやりかたを教えてくれました。

# 既存の設定を削除
sudo tc qdisc del dev eth0 root (何か表示されるが無視する)

# `netem` を root にして、まずパケットロスを設定
sudo tc qdisc add dev eth0 root handle 1: netem loss 1% delay 10000ms

# その上に `tbf` を適用
sudo tc qdisc add dev eth0 parent 1:1 handle 10: tbf rate 10mbit burst 10kb latency 500ms

つまるところ、ルールの親と子を引っくり返す、という手法です。

まあ、理由は不明ですが、これで動いているようです。

2025,江端さんの忘備録

『DeepSeekは、本当に"オープンソース"なんですか?』とChatGPTに尋ねてみました。

I asked ChatGPT, “Is DeepSeek really ‘open source’?”

===== 解答ここから =====

===== Answer from here =====

はい、DeepSeekはオープンソースのAIモデルを開発・公開しています。同社のAIモデルは、コードやアルゴリズムが公開されており、誰でも自由に利用・改良が可能です。

Yes, DeepSeek develops and publishes open-source AI models. The company's AI models are publicly available code and algorithms; anyone can use and improve them.

例えば、DeepSeekがリリースした「DeepSeek-R1」モデルは、オープンソースとして提供されており、開発者が自由に利用・改良できるようになっています。

For example, the DeepSeek-R1 model released by DeepSeek is open source, allowing developers to use and improve it freely.

このようなオープンソース戦略により、DeepSeekはAIコミュニティ内での信頼性を高め、技術の透明性を確保しています。

This open-source strategy gives DeepSeek credibility within the AI community and ensures technology transparency.

===== 解答ここまで =====

===== Answer to here =====

これ以上聞いたら、ChatGPTが私に腹を立ててサービスを停止するかも、という、(完全にデマレベルの)恐怖を感じて、これ以上の質問は止めました ―― 今、ChatGPTが止まったら、私の仕事も学業も止まってしまうので。

I felt a (totally groundless) fear that if I asked any more questions, ChatGPT might get angry with me and stop providing its service, so I stopped asking any more questions. If ChatGPT stopped now, my work and studies would also stop.

それはさておき、現時点で私の拾った情報を記載します。多分、嘘やデマやデタラメが入っているはずなので、基本的には以下の内容は、そのまま信じずに自分で調査して下さい(私も調べます)。

Leaving that aside, I will now describe the information I have gathered. There is probably some false information, rumors, and nonsense in there, so please don't believe the following content as it is and research it yourself (I will investigate it, too).

(1)オープンソースである → 原則として誰もが自由につかえて、改修も可能である → 私が個人的に「生成AI」を作れる可能性がでてきた

(1) It is open source → In principle, anyone can use it freely and make modifications → The possibility of me being able to create “generative AI” has arisen.

(2)ChatGPTと比較して、1/10のコストで開発 → 金はないが、時間さえかければ、私個人でもなんとかなるかもしれない

(2) Compared to ChatGPT, it can be developed at 1/10 the cost → I don't have any money, but if I put in the time, I might be able to manage it myself.

(3)学習に高性能、高機能のデバイスを必要としない → ゲーミングPCレベルのスペックがスコープに入ってくるかもしれない

(3) I don't need a high-performance, high-function device for learning → I might be able to get by with specs on the level of a gaming PC.

私も、これらの全部を、すんなりと信用することはできませんが、真偽のほどは、いずれ分かるでしょう。

I cannot trust these things easily either, but the truth will be known eventually.

-----

米国のメディアは「スプートニクショックと同程度の衝撃」という表現で、DeepSeekを表しています。

The American media described DeepSeek as “having the same impact as the Sputnik shock.”

これ、私は「大袈裟」な表現とは思っていません。

I don't think this is an “exaggerated” expression.

ただ、私は「スプートニクショック」を知らないので、「Linuxカーネル(オープンソース & フリー)登場時と同程度の衝撃」と表現したいと思います。

However, since I don't know about the “Sputnik Shock,” I would like to express it as “having the same impact as the appearance of the Linux kernel (open source & free).”

私の趣味を、業務が奪っていく ――

2025,江端さんの忘備録

江端:「将来、娘たちをストーカーする男を殺害して、首や足や腕を裁断する時には、もっと血が出るんだぞ。分解した体を部位をコンクリーで固めて、最後に、東京湾に沈めなければならないのに、そんなことで大丈夫なのか?」

にも記載していますが、江端家では、正月には鰤(ぶり)の一本買いをしています。

As mentioned above, the Ebata family buys a yellowtail for New Year's.

ぶりを一本買いをすると、「ぶりの刺身の『永久機関』」というくらい、毎日(成人の日くらいまでは)ぶりの刺身を楽しめます(しかも、スーパーの切り身を購入するのが、バカバカしいくらい安いコストで)。

If you buy a whole yellowtail, you can enjoy sashimi of yellowtail every day (until about Coming of Age Day).

私のような魚をおろすのがヘタクソな素人であっても、身が残った骨と皮を、そのまま大きなフライパンに放り込んで、塩とこしょうでソテーするだけで、(見栄えはどうあれ)無駄なく、全部美味しく頂けます ―― 使用率100%です。

Even if you're a novice like me who's terrible at filleting fish, you can throw the bones and skin with the flesh on them straight into a large frying pan and sauté them with salt and pepper. Then, you can enjoy all of it without wasting it (regardless of how it looks)—100% utilization.

-----

先日、ある番組で、この「ぶりを丸ごと一本送る風習がある」と知りました。

The other day, I learned on a TV program that the custom is to send a whole yellowtail.

■嫁ぶり:結婚した年の年末に婿側の実家から嫁側の実家にブリを贈る風習

Bride's behavior: A custom in which the bride's family receives a yellowtail from the groom's family at the end of the year in the wedding year.

九州北部や岡山県、富山県西部などで行われているらしいのですが、私は、この風習を全く知りませんでした。

It seems to be practiced in northern Kyushu, Okayama Prefecture, western Toyama Prefecture, etc., but I had never heard of this custom before.

で、その際、ぶりは「大きければ大きいほど良い」とされているようで、13kgのぶりを送ることもあるそうです。

Apparently, the bigger the better when it comes to yellowtail, and they sometimes send yellowtail that weighs 13kg.

ちなみに、上記の写真のぶりでも4kgクラスです。

Incidentally, the yellowtail in the photo above is also in the 4kg class.

―― 13kgのぶりって、一体どうするんだ?

"What on earth do they do with a 13kg yellowtail?"

と思いましたよ。まあ、ご近所や親戚に配るらしいですが、私は『面倒くさいなぁ』と思ってしまいました。

They seem to give them to neighbors and relatives, but I thought, “What a hassle.”

-----

昨年末のぶりの値段は、従来の2倍まで値上がりしていました。

The price at the end of last year had risen to twice the previous price.

水揚げが減ったという話は聞いていなかったので、変だな、と思っていたのですが、どうやら、米国が大量買いをしていたらしいです。

I hadn't heard that the catch had decreased, so I thought it was strange, but the US was buying a significant amount.

この魚の美味しさに、世界が気がつきはじめています。ヤバイです。

The world is starting to notice how delicious this fish is. It's dangerous.

大晦日前にジャストインタイムで届くぶりは、値段が高くなります。そこで、大変不本意ではありましたが、私は、一週間も前に、ぶりを購入し、解体して、冷凍庫に突っ込んでおきました。

The price of a yellowtail that arrives just before New Year's Eve is higher. Although very reluctant, I bought the yellowtail a week in advance, gutted it, and put it in the freezer.

まさか、クリスマスイブの日に、血だらけになりながら魚の解体をやることになるとは思いませんでしたが。

I never thought I'd be covered in blood on Christmas Eve for gutting fish.

あ、それと、素人の大型鮮魚の解体は、玄人のようにはいきません ―― 今回、私、『のこぎり』つかって、ぶりの頭を切り落していました。

Oh, amateur fish butchers can't do it as well as the professionals - this time, I used a saw to cut off the head of the yellowtail.

あれ、解体作業がラクになるのでお勧めです ―― 見た目の美しさを無視すれば、ラクできますよ。

It makes the dismantling process easier, so I recommend it. It will be easier if you ignore the visual appearance.

『これは、娘のストーカーの死体の解体の際にも有用だな』と思いながら、のこぎりを振り回していました。

I was swinging the saw around, thinking, 'This will be useful when I dismember my daughter's stalker's corpse.

-----

江端家では、刺身作りの作業を、毎回、私がやらされているので、

At the Ebata house, I am made to do the sashimi preparation every time, so

『私が死んだ後は、どうするんだ!』

"What are you going to do after I die!"

と一喝したところ、

I shouted,

『パパが死んだ後、誰が、ぶりの一本買いをするの?』

'After you die, who's going to buy a whole fish?

と問われて、私は沈黙せざるを得ませんでした。

I had no choice but to remain silent.

未分類

tc コマンドの帯域制限 (tbfnetem) は 送信(アウトバウンド, egress) に適用されます。したがって、上記のコマンドは eth0 から外部に送信されるトラフィック に対して帯域制限が適用されます。


eth1, eth2 の場合の制限方向

eth1eth2 がある場合、それぞれのインターフェースに tc を適用すると、次のようになります。

  • tc qdisc add dev eth1 root tbf rate 5mbit burst 10kb latency 1ms
    • eth1 から外部へ送信するトラフィック(egress) に 5Mbps の帯域制限
  • tc qdisc add dev eth2 root tbf rate 5mbit burst 10kb latency 1ms
    • eth2 から外部へ送信するトラフィック(egress) に 5Mbps の帯域制限

つまり、各インターフェースから送信される通信に適用される ため、受信(ingress)には影響しません。


受信(ingress)側の帯域制限をしたい場合

tc は基本的に送信 (egress) に適用されるため、受信 (ingress) の帯域制限を行うには ingress qdisc を利用するか、ifb (Intermediate Functional Block device) を使います。

eth1 の受信帯域を 5Mbps に制限する場合

sh
# ifb デバイスを有効化
sudo modprobe ifb
sudo ip link set ifb0 up
# eth1 の受信トラフィックを ifb0 にリダイレクト
sudo tc qdisc add dev eth1 handle ffff: ingress
sudo tc filter add dev eth1 parent ffff: protocol all u32 match u32 0 0 action mirred egress redirect dev ifb0
# ifb0 に帯域制限を適用(受信制限)
sudo tc qdisc add dev ifb0 root tbf rate 5mbit burst 10kb latency 1ms

これにより、eth1 への受信トラフィック を 5Mbps に制限できます。


まとめ

  • tc のデフォルトの動作では 送信(egress) のみ帯域制限される。
  • tceth1 に適用すると、eth1 から出ていく通信に制限がかかる。
  • 受信(ingress)を制限したい場合は ifb デバイスを使う。

したがって、最初のコマンドは eth0 の送信(egress)方向 に適用される帯域制限になります。

 

 

2025,江端さんの技術メモ

例題: 水道の蛇口に例える

ネットワークの「ビットレート (bitrate)」と「帯域 (bandwidth)」 の違いを、水道の蛇口に例えて説明します。


1. 帯域 (bandwidth) = 水道管の太さ

  • 水道管の太さが 「ネットワークの帯域」 に相当する。
  • 例えば 「10Gbpsの回線」 というのは 「直径10cmの水道管」 に相当する。
  • 「最大で10Gbps流せる」 という意味だが、実際にどれくらい流すかは別の話。

2. ビットレート (bitrate) = 水の流量 (蛇口の開き具合)

  • 蛇口をどれくらい開くかが 「ビットレート」 に相当する。
  • 例えば tcrate 5mbit を設定すると、蛇口を半分だけ開くようなイメージ。
  • 回線が10Gbps(帯域)でも、実際に使うのは5Mbps(ビットレート)に制限される。

3. tc を使った制限の具体例

(1) 10Gbpsの回線を持つPC (帯域 = 10Gbps)

  • 例えば、自宅のネット回線が 10Gbps だったとする。
  • しかし、tc を使って rate 5mbit に制限すると、実際のデータ転送は 5Mbps しか出ない。

(2) tc で 5Mbps のビットレート制限

sudo tc qdisc add dev eth0 root tbf rate 5mbit burst 10kb latency 1ms

この設定をすると、10Gbpsの回線を持っていても、実際の転送速度は 5Mbps に制限 される(=蛇口を少ししか開かない)。

(3) tc で 5Mbps ~ 10Mbps の変動を許可

sudo tc qdisc add dev eth0 root handle 1: htb default 10
sudo tc class add dev eth0 parent 1: classid 1:1 htb rate 5mbit ceil 10mbit
  • 最低 5Mbps最大 10Mbps まで許可 する設定。
  • 普段は5Mbpsだが、ネットワークに余裕があれば 最大10Mbps まで利用可能

4. 結論

項目 水道の例 ネットワークの意味
帯域 (bandwidth) 水道管の太さ 利用可能なネットワークの最大容量 (例: 10Gbps)
ビットレート (bitrate) 蛇口の開き具合 実際に流せるデータ速度 (例: 5Mbps)
  • 帯域は「最大でどれくらい流せるか」 を決める(物理的な上限)。
  • ビットレートは「実際にどれくらい流すか」tc で調整できる。

2025,江端さんの忘備録

先日、こんな日記を投稿しましたが、

I posted a diary like this the other day.

■若手のエネルギッシュな提案は――うっとうしい。

これ、実は若いころの私のことです。

The person in this diary was me when I was young.

まあ、本当に色々やりましたよ。

Well, I did a lot of things.

■自作の特許明細書執筆フォローシステム作って運用したり、

I created and operated a system for writing and following up on patent specifications. 

■研究所内に監視カメラシステムを導入したり、

I had installed a surveillance camera system in the laboratory,

■勉強会を立ち上げたり、

I had started study groups, etc.

■組合に「バリケードストライキ」を提案したり、

I had proposed a “barricade strike” to the union,

■頼まれれば、どの国にもホイホイと海外出張して、デタラメ英語で打ち合わせしてきたり、

If asked, I had happily traveled abroad to any country and attended meetings in my made-up English.

今から考えると、

Looking back now,

―― 背筋も凍るほど『鬱陶(うっとう)しい若手研究者』

"I was a 'young researcher who was so annoying that it made my spine tingle."

でした。

さぞ、うっとうしい奴だったでしょう。

I must have been a very annoying person.

あまりにもうっとうし過ぎて、私に殺意を覚えたシニア研究員もいたと思います ―― いや、絶対にいたはず。

I think there was a senior researcher who was so annoyed that they felt like killing me - no, there was.

今の私が、本気で"そう"感じているのですから。

Because I think that way now.

-----

でも、当時、そんな「うっとうしい私」に文句も言わずに、みんな我慢してくれました。

But at the time, everyone put up with my annoying ways without complaining.

だから、今度は、私が「我慢する側」にならなければならないのです。

So, this time, I have to be the one who “puts up with it.”

『若手のエネルギッシュな提案』は、絶対に潰してはなりません。絶対に、です。

'The energetic proposals of the young' must never be squashed. Never.

可能であれば、提案や助言もしない方が良いでしょう(私も、言いたいことは沢山あるんですけど、歯をくいしばって「沈黙」をキープしています)。

If possible, it's probably best not to make any suggestions or give any advice (I have a lot I want to say, but I'm gritting my teeth and keeping quiet).

私たちシニアができる、かれらに対する最大の貢献は、『無干渉(×無関心)』です。

The most outstanding contribution we seniors can make to them is to 'stay out of their way (and not be indifferent to them).'

「やりたいことを、やりたいように、やらせてあげる」が、私たちシニアの最後の努めです ―― たとえ、それが失敗に向かっていることが、明白であったとしてもです。

Our last duty as seniors is to “Let them do what they want to do. However, they want to do it,” even if it is clear that they are heading for failure.

私たちは、だれであれ「失敗からしか学ぶことができない」からです(単に、私の個人の信念ですが)。

Because we can only learn from our mistakes (this is my personal belief).

『"失敗"からしか学べないなら、学ぶためには"失敗"し続けなければならない』

2025,江端さんの忘備録

南海トラフ地震は、史実で残っているものだけでも、9回観測されているそうです(NHKの『歴史探偵』)。

It is said that the Nankai Trough Earthquake has been observed 9 times in history alone (NHK's “Rekishi Tantei”).

先日、今度の南海トラフ地震の発生率が『30年以内の発生率80%』と上方修正されました。

The other day, the probability of a Nankai Trough earthquake occurring within the next 30 years was revised upwards to 80%.

地球が地殻変動を続ける限り、地震のエネルギーは毎日蓄積され続けているわけですから、地震の発生を止める手段はありません。

As long as the earth continues to experience crustal movement, the energy for earthquakes will accumulate daily, so there is no way to stop them from occurring.

で、ふと思ったのですが、

Then I suddenly wondered,

―― 南海トラフ地震を、人為的に発生させることってできないかな?

“Can't we cause the Nankai Trough earthquake artificially?”

と。

人為的であれ、南海トラフ地震を発生させれば、大変な被害が予想されますが、日時が決っていれば、震災による被害を最小化できることは確かです(まあ、地震や津波による家屋倒壊や、地価の変動や市場の混乱は仕方がない、と腹を括ります)。

Even if it is artificial, if the Nankai Trough earthquake occurs, it is expected to cause significant damage. Still, suppose the date and time are known. In that case, the damage caused by the earthquake can inevitably be minimized (well, we will have to accept that there is nothing we can do about the collapse of houses due to the earthquake and tsunami or the fluctuations in land prices and market confusion).

これによって、人類史上、始めての『震災死ゼロ』が実現できるかもしれません。

This may be the first time in human history that there have been no deaths from an earthquake.

-----

ただ、私、この辺について、故小松左京先生の「日本沈没」で出てきた地震のエネルギー量をざっくりと覚えており、地震を誘発するのに、水素爆弾程度のエネルギーでは小さすぎるだろうなぁ、とも思っていました。

However, I vaguely remember the amount of energy released by the earthquake in the novel “Japan Sinks” by the late author Sakyo Komatsu. I also thought that the energy released by a hydrogen bomb would be too small to trigger an earthquake.

ChatGPTに聞いたところ、まあ、同じような解答を得ました。

When I asked ChatGPT, I got a similar answer.

内容や言葉は丁寧でしたが、『寝惚けたこと言ってんじゃねーよ』という内容でした。

The content and words were polite, but the content was, 'Don't say stupid things like that when you're half asleep.

===== ChatGPT解答ここから =====

===== ChatGPT Answer from here =====

(前略)

(omission)

結論

Conclusion

南海トラフ地震を水素爆弾で発生させることは、エネルギー的に見ても技術的に見てもほぼ不可能です。仮に何らかの方法で誘発できたとしても、予測不能な大災害を引き起こす可能性が高く、計画的にコントロールすることは現時点では不可能です。

It is almost impossible to cause a Nankai Trough earthquake with a hydrogen bomb in terms of energy and technology. Even if it were somehow possible to induce one, there is a high possibility of causing an unpredictable disaster, and it is currently impossible to control it systematically.

むしろ、現実的な対策としては早期警戒システムの強化や、建築基準の向上、防災対策の充実を進める方が、遥かに合理的なアプローチといえます。

A more rational approach would be strengthening early warning systems, improving building standards, and enhancing disaster prevention measures.

===== ChatGPT解答ここまで =====

===== ChatGPT Answer ends here =====

毎日100~200の地震が発生しているのですから、「大地震が起きませんように(もし起きるなら、私の関係のない場所でヨロシク)」と祈っても、多分、無駄でしょう。

音声AI:「だからぁ! 津波警報が発令されてヤバいんだってば! 死にたくなければ、直ぐに逃げろ!! 40秒で支度しな!!」

未分類

「『もうすぐやってくるぞ』詐欺」というもの一つとして、「シンギュラリティ」というものがあります。

One example of the “It's coming soon” scam is the concept of “singularity.”

この「シンギュラリティ」って、面白いんですよ。

This “singularity” is interesting.

第一に、「シンギュラリティ」なる"モノ"が数値で定義されておらず、未だに概念のみの存在であること。

Firstly, the concept of “singularity” is not defined by a numerical value and still only exists as a concept.

第二に、その発生予想時期が「現在から数年または数十年後」に、常に更新され続けていること。

Secondly, the expected time of occurrence is constantly being updated to “a few years or decades from now.”

まさに「永遠のシンギュラリティ」です。

It is truly the “Eternal Singularity”.

(↑コラムに飛びます)

(Jump to the column)

現在の私が、生成AI(ChatGPT等)を使い倒していることはご存知の通りです。

As you know, I am using generative AI (such as ChatGPT) to the fullest.

そして、生成AIが、完全に無知性であり、文章の作り方(単語の選択、文字の並び、文章の心地良さ)だけを、計算した結果を表示しているものであることは、こちらで説明しています。

I explain here that the generated AI is completely unintelligent and only displays the results of calculations on how to make sentences (choosing words, arranging letters, and determining how pleasant the sentence is).

(↑コラムに飛びます)

(Jump to the column)

人間の心とか知性とか呼ばれるものは、「しょせん、機械によって作らされる程度のものである」と、私は、人間の能力を『下方修正』しています。

I am 'downwardly revising' human abilities, saying that the human mind or intellect is “only something that is created by machines.”

私たち人間が考える「心」とか「知性」とか言うものは、単語の選択、文字の並び、文章の心地良さ程度の"モノ"や"コト"で、その大部分が決まる、といってもいいんじゃないか、と思うんです。

I think that the “mind” and “intelligence” that humans feel about are primarily determined by word choice, the letters' order, and the sentence's pleasantness.

この辺りについては、すでに、これまでも論じてきた通りです。

I have already discussed this area.

―― 私たちは、心地の良いコミュニケーションだけがあればいい

つまり、「『シンギュラリティ』なんて言葉は、もういらない」ということです。

In other words, “we don't need the word ‘singularity’ anymore.”

-----

「創作だけは機械にできない」と言われていますが、私がChatGPTを使い倒している日々において、私は、"これ"すらも"否定"できます。

It is said that “only creative work cannot be done by a machine,” but in my daily use of ChatGPT, I can even “deny” this.

多くの人は、「創作」というのが、「まっしろなキャンパスの中から、突然湧き上がる発想」みたいな考え方をしているかもしれませんが ―― これ、素人の考え方です。

Many people may think that “creation” is something like “an idea that suddenly springs up from a blank canvas,” - but this is the way of thinking of amateurs.

私のような、「特許明細書製造マシン」から見れば、「創作」とは、『既往技術の組み合わせによって、単なる組み合わせ以上の価値を生じさせるモノまたはコト』です。

From the perspective of a “patent specification manufacturing machine” like me, “creation” is “something or something that creates value beyond mere combination by combining existing technologies.”

つまり、発明とは「既存技術の足し算」と「その足し算以上の価値の生成」です。

In other words, the invention is “adding to existing technology” and “creating value greater than the sum of the additions.”

そして、ChatGPTは、この「組み合わせ」がビックリするくらい上手い。

And ChatGPT is surprisingly good at this “combination.”

「どうやって、その組み合わせを思いついた?」と尋ねたくなるくらいです。

I can't help but ask, “How did you come up with that combination?”

-----

例えば、私、昨夜、「人の交流度を上げる式の定式化の手法として、交流する場所の『粘性(動きにくさ)』を入れることができるか?」と、ChatGPTに尋ねたら、以下のように解答されて、度肝を抜かれました。

For example, last night, I asked ChatGPT, “Can the ‘viscosity (difficulty of movement)’ of the place of interaction be included in a formula to increase the degree of human interaction? "The following answer blew me away.

=====

「交流する場所の粘性が高いと親密度が高まる」という仮説を表現するのに適した物理法則として、以下が挙げられます。

The following are suitable physical laws for expressing the hypothesis that “the viscosity of a place for interaction increases intimacy.”

  1. ハーゲン・ポアズイユの法則(粘性が高いと流動性が低下 → 人の移動が減り、親密度が増加)
    Hagen-Poiseuille's law (when viscosity is high, fluidity decreases → human movement decreases, intimacy increases)
  2. 万有引力の法則(移動が制限されることで距離が縮まり、相互作用が強くなる)
    Universal Law of Gravitation (distance is shortened, and interaction becomes stronger when movement is restricted)
  3. ボルツマン因子(粘性が高くても、ストレスが大きいと親密度が下がる影響を考慮)
    Boltzmann factor (even if viscosity is high, the effect of a large amount of stress lowering intimacy is taken into account)
  4. 粘性応力の関係(粘性が高いと流速勾配が小さくなり、固定的な関係が強まる)
    Viscous stress relationship (when viscosity is high, the flow velocity gradient decreases, and the fixed relationship becomes stronger)

最も単純なモデルとしては、「親密度 は粘性 に比例する」という関係が適用でき、ハーゲン・ポアズイユの法則や万有引力の法則の変形で表現できる。

The simplest model is one in which “proximity I am proportional to viscosity η” can be expressed as a variant of the Hagen-Poiseuille law or the law of universal gravitation.

(以下、定式モデルの記述については、割愛)

(The following description of the formula model is omitted.)

=====

で、昨夜遅くまでChatGPTと議論して、この"粘性"について一般式に落し込むところまで完了しました。

So, I was discussing with ChatGPT until late last night, and we completed reducing this “viscosity” to a general formula.

この作業は『ChatGPTと一緒に「創作」をした』と断言できる内容です。

This content can be asserted as “creative work” done with ChatGPT.

少なくとも上記の私の「創作」の定義は、100%満たしています。

At the very least, my definition of “creation” above is 100% fulfilled.

-----

私たちは、シンギュラリティなど待っている必要はありません。

We don't need to wait for things like Singularity.

今、私たちに必要なのものは、生成AIの「使い倒し方」(あるいは、「引き出しの見付け方」)の一択だと思います。

I think that what we need right now is a way to “get the most out of” (or “find the best way to use”) generative AI.

まあ、正直に言えば、私は、この「使い倒し方」を、できるだけ長い間、誰にも知られたくないですが。

I want to keep the max” secret for as long as possible.

「まだChatGPTを使ってない人は『人生を悔い改めた方がいい』」 ―― と言った、ソフトバンクの孫社長に申し上げます。『いらんこと言うな』と。

2025,江端さんの忘備録

私は、自分のために、自分のやったことをメモに残すことを常としています。

I always make a note of what I've done for myself.

で、秘密情報に当たらないものについては公開もしています。

And I also make public information that does not fall under the category of confidential information.

さくらインターネット上のWordPressをエクスポートしたものを、Amazon Lightsailにインストールする方法

なので、これについても、メモを残そうと思っていたのですが ―― 断念しました。

So I was going to write a memo about this too, but I gave up.

とにかく、思うように移行作業が進まなくて、ChatGPTと協議しながらの場当たり対応にならざるを得ませんでした。

Anyway, the migration process didn't go as smoothly as we had hoped, and we had to respond on an ad hoc basis while consulting with ChatGPT.

移行作業のためだけに、別のサーバを立ち上げて、そのサーバを動かすために設定ファイルを何十回も書き直し(文字通り、"何十回")しました。

I had to set up a separate server just for the migration process, and I had to rewrite the configuration file dozens of times (literally, “dozens of times”) to get it working.

そうしているうちに、ChatGPTの指示が同じ内容を繰り返すようになり ―― つまり、ChatGPTもアドバイスできることがなくなり ―― 完全にスタックしてしまいました。

While doing this, ChatGPT's instructions started to repeat the same content - in other words, ChatGPT also stopped being able to give advice - and it completely got stuck.

ユーザを変更する為だけに、使ったことのないデータベースのコマンドを打ち込み、エラーが消える外観のテーマ(WordPressを使っている人なら分かる)を選ぶという低落ぶりです。

It's so bad that I had to type in commands for a database I'd never used before just to change the user, and then choose a theme that made the errors disappear (if you use WordPress, you'll know what I mean).

最大の問題は、移行前のサイトと移行後のサイトのバージョン違いと、そしてWordPressの(一方的な)仕様変更です。

The biggest problem is the difference in versions between the site before the migration and the site after the migration, and the (one-sided) specification changes in WordPress.

『なんで開発サイトは、そんなに仕様変更したいのかな』と思うのですが、まあ、その理由は、セキュリティ対策であることは明らかです。

I wonder why the development site wants to change the specifications so much, but the reason is obviously for security measures.

セキュリティホールを見つけると、『鬼の首を取ったかのように、それを喧騒する奴(技術サイト)』が、世界中には山ほどあります。

There are tons of websites around the world that “make a big deal out of it as if they've taken the head of a demon” when they find a security hole.

対策を怠ると、世界中のWordPressの利用者の情報が筒抜けになる ―― と考えれば、WordPress開発サイドの、この対応はやむを得ないとも思います。

If they neglect to take measures, the information of WordPress users around the world will be exposed - if they think about it like that, I think that the response from the WordPress development side is unavoidable.

-----

『だから、秘密情報(パスワード、銀行口座番号など)は、ノートだけ記載して、それを分からない場所で管理しておくべきなんだよ』

'That's why you should only write down secret information (passwords, bank account numbers, etc.) in a notebook and keep it in a place where it can't be seen.

という、嫁さんの『アナログ対応戦略』こそが、正解だったことを、今、実感として感じています。

I now realize that my wife's “analog strategy” was the correct one.

-----

嫁さん:『秘密情報は全部手書きで書き出してしまえば?』

Wife: “Why don't you write all the confidential information out by hand?”

私:『死ぬまでに、全てを書き出しが終えれるか分からん。多分、無理だと思う』

Me: 'I don't know if I'll be able to finish writing everything before I die. I probably won't be able to do it.

-----

私の持っている秘密情報とは、エロいやつとか、そういうレベルのものではありません(そんなもの、知りたい人がいれば、いくらでも教えてあげます)。

The secret information I have is not of the pornographic kind (if anyone is interested in that kind of thing, I'll be happy to tell them).

そして、手書きできるようなレベルのものでもありません。多くは語れませんが、結構ヤバいものもあります。

And it's not even at a level where I can write by hand. I can't say much, but there are some pretty bad ones.

もちろん、それが開示されたところで、直ちにバレるような内容とはなっていません。

Of course, the information disclosed does not contain anything that would be immediately recognizable.

ただし、そこに『江端の拉致と拷問』という手段で『秘密鍵』を入手すれば、『開錠』することができる ―― という"モノ"です。

However, if you obtain the “secret key” by means of “Ebata's abduction and torture”, you will be able to “unlock” it.

例えば、『コラムのページの中に書き込まれている一行のパスフレーズを、パスワードに変更できるのは私だけ』という感じの"モノ"です。

For example, it is something like “Only I can change the one-line passphrase written in the column page into a password”.