ChatGPTとGoogle Bardに”mix”について聞いてみた

雑記

ChatGPTとGoogle Bardに"mix"について聞いてみた

どうも。
日本のアイドルライブではmixと呼ばれるコール?声援?があります。
これについてGPT(4)とGoogleのBardに聞いてみたので共有します。

質問1:mixを知ってるか?

Do you know the call made at Japanese idol concerts called “mix"?
(日本のアイドルのコンサートで行われる「mix」というコールをご存知ですか?)

とききました(Bardはまだ日本語未対応なので

結果

回答

Bard

Yes, I am familiar with the “mix" call that is made at Japanese idol concerts. It is a chant that is typically shouted during the introduction and instrumental breaks of an idol’s song. The mix is universal and is not associated with a specific idol group. It consists of three main variations: the Standard Mix, the Japanese Mix, and the Ainu Mix. Some idols also have their own specific versions of the Mix unique to them.

  • The Standard Mix is the most commonly used mix of the three, and is also known as the First Mix as it is typically heard during the introduction of the song. In English, it follows the order below.
Tiger, Fire, Cyber, Fiber, Diver, Viber, Jyaa jyaa!!
  • The Japanese Mix is the translated Standard Mix, and is typically heard at the end of the first instrumental break, leading into the second verse of the song. Due to this, the Japanese Mix is sometimes referred to as the Second Mix.
Ah, Mou iccho ikuzo! Tora! Hi! Jinzou! Seni! Ama! Shindou! Kasen! Tobi! Joukyo!
  • The Ainu Mix is a more recent variation of the mix that is inspired by the traditional music of the Ainu people of Japan. It is typically used during songs that have an Ainu feel to them.
Ainu, Ainu, Ainu, Ainu, Ainu, Ainu!

The mix is a way for fans to show their support for their favorite idols and to create a more energetic and exciting atmosphere at concerts. It is also a way for fans to connect with each other and to create a sense of community.

そう、日本のアイドルのコンサートで行われる「ミックス」コールを私は知っている。これは、一般的にアイドルの曲のイントロや楽器のブレイク時に叫ばれるチャントである。ミックスは普遍的なもので、特定のアイドルグループとは関係ありません。スタンダード・ミックス、ジャパニーズ・ミックス、アイヌ・ミックスの3つのバリエーションから構成されています。また、一部のアイドルは、そのアイドル独自のミックスバージョンを持っています。

スタンダードミックスは、3つのミックスの中で最も一般的に使われているミックスで、曲のイントロで聴かれることが多いため、ファーストミックスとも呼ばれる。英語では、以下のような順番になります。
Tiger、Fire、Cyber、Fiber、Diver、Viber、Jyaa jyaa!!
Japanese Mixは、Standard Mixを翻訳したもので、通常、最初のインストゥルメンタル・ブレイクの終わりで、曲の2番目のバースにつながるところで聞かれます。このため、Japanese MixはSecond Mixと呼ばれることもあります。
ああ、モウイッチョイクゾウ!寅さん!ハイ!ジンゾウ!セニ!あま!新堂!カセン!トビ Joukyoです!
アイヌミックスは、日本のアイヌ民族の伝統音楽にインスパイアされた、より新しいミックスのバリエーションです。一般的には、アイヌの雰囲気が漂う曲で使われることが多い。
アイヌ、アイヌ、アイヌ、アイヌ、アイヌ!
このミックスは、ファンが好きなアイドルを応援する気持ちを表し、コンサートでよりエネルギッシュでエキサイティングな雰囲気を作り出すための方法である。また、ファン同士がつながり、コミュニティーの感覚を生み出す方法でもあります。

www.DeepL.com/Translator(無料版)で翻訳しました。

GPT4

Yes, I’m familiar with the “mix" call at Japanese idol concerts. A “mix" (ミックス or mikkusu) is a form of audience participation, which involves coordinated chants and cheering, accompanied by rhythmic clapping or waving lightsticks. This energetic display is an important part of the Japanese idol culture and is intended to show support and enthusiasm for the performers.

The mix call usually occurs during instrumental breaks or during the chorus of a song. Fans, often called “wota" (a term derived from “otaku," meaning a passionate fan), will shout a series of words or phrases in a specific order, creating a sense of unity and energy in the crowd. Common phrases used in a mix call include the names of the idol group members, general words of encouragement, or even nonsensical phrases that have become associated with the particular idol group.

While mix calls can vary depending on the group and the song, there is often a standard pattern that fans follow. It is not uncommon for fans to practice these chants and coordinate their efforts ahead of time to create an impressive and synchronized display of support during the concert.

そう、日本のアイドルのコンサートでの「ミックス」コールを知っている。ミックス」とは、観客の参加型演出の一つで、リズミカルな手拍子やライトスティックを振って、チャントや声援を協調的に行うものです。このエネルギッシュな演出は、日本のアイドル文化の重要な一部であり、パフォーマーへのサポートと熱意を示すことを目的としています。

ミックスコールは通常、楽器の休憩時間や曲のコーラス中に行われます。ヲタ(熱狂的なファンを意味する「ヲタク」から派生した言葉)と呼ばれるファンは、特定の順序で一連の単語やフレーズを叫び、観客に一体感とエネルギーを与えます。ミックスコールでよく使われるフレーズは、アイドルグループのメンバーの名前、一般的な励ましの言葉、あるいは特定のアイドルグループから連想されるナンセンスなフレーズなどである。

ミックスコールは、グループや曲によって様々ですが、ファンが行う定番のパターンがあることが多いようです。また、事前に練習を重ね、コンサート本番で見事なシンクロを見せることも珍しくありません。

www.DeepL.com/Translator(無料版)で翻訳しました。

考察

圧倒的にBardの方が正しいといえる。アイヌはしらんが、、、
○○教えてっていう質問は基本的に事実を回答するだけなので検索結果を提示するだけでいいのですが、
GPTは学習データをサーバーに保管していて、そこにあるデータを参照して答えを出していると思われる。

あなたは知らないことを聞かれたらどのように回答を生成していますか?

私はAIモデルであり、あなたが質問する内容が私の学習データ内に存在しない場合、私が知らないことに対して回答を生成することは困難です。ただし、私は以下のようなアプローチで回答を生成しようとします。

  1. 類似性のある情報を利用:私は学習データ内にある似たような情報を使って、質問に対する回答を推測します。これにより、完全に正確でないかもしれませんが、あなたが求めている情報に近い回答を提供できる可能性があります。
  2. 論理的推論:私は論理的思考を使って、既存の知識から新しい情報を導き出そうとします。これにより、質問に対して一般的な答えや関連する情報を提供できることがあります。
  3. 問題の再定義:私は質問をより具体的で解答可能な形に再定義し、その回答を提供することがあります。これにより、私が知っている範囲内で質問にできるだけ近い回答を提供することができます。
  4. 知識不足を認める:最後に、私は知識が不足していることを認め、その情報を提供できないことを示すことがあります。私は常に正確な情報を提供しようと努力していますが、私の学習データには限界がありますので、すべての質問に答えることはできません。

ので日本のごく一部で使われている謎の絶叫については学習しておらず、知ってる情報から推測して回答したけっかこれであると思われる。

のでGPTを利用し、bingという検索エンジンに組み込まれたchatでは最低限の回答ができていた。

Yes, I know about the call made at Japanese idol concerts called “mix”. It is a type of call that fans use to cheer on their favorite idols during live performances. There are many different types of mix calls, such as standard mix, encho mix, kahen mix, Japanese mix, etc 1. Would you like to know more about it?
詳細情報:

  1. idolcall.jp
  2. idolcall.jp
  3. idolcall.jp
俺のブログを参照しろよ…

それと比較してBardはおそらくBingと一緒で知らないことはリアルタイムでググってる。
○○教えて系については圧倒的にリアルタイムでググって要約した方が強い。

質問2:mixを考えろ

実際に考えさせます。
GPTにはbardの情報をおしえて考えさせます。(アイヌは直しました)

回答

bard

Sure, here is a new “mix" that I created based on the above information:

**Fire, Ice, Wind, Earth, Light, Dark, Love, Hate!**

This mix is inspired by the elements of nature, as well as the forces of good and evil. It is a call to action for fans to come together and create a more positive and harmonious world.

I hope you like it!

GPT4

Certainly! Based on the structure and style of the existing mixes, I’ve created a new “mix" variation for you. This mix is inspired by elements of nature and incorporates both English and Japanese words:

Ocean, Earth, Kaze (風), Hikari (光), Forest, Umi (海), Thunder, Yama (山), Sora (空), Freedom!

This mix translates to:

Ocean, Earth, Wind, Light, Forest, Sea, Thunder, Mountain, Sky, Freedom!

Remember, the purpose of a mix is to create a sense of unity, energy, and support for the idols during their performances. This new mix variation can be used during introductions or instrumental breaks, just like the existing mixes. Feel free to share it with fellow fans and enjoy the exciting atmosphere it brings to concerts.

考察

日本のキモオタクが考えるようなクソmixはAIには考えられないっぽいですね。
一応mixっぽいものにはなっていますが、面白いかって言われると、今更このようなmixをするオタクはいないでしょう。

質問3:mixを考えろ

もう少し条件を絞ります。

Please consider a new “mix" with the following conditions
A humorous mix that will make people laugh.
Use 16 bars (the length of 8 standard mixes)
No specific theme, but consistency would be nice.
Easy to remember
Use phrases that can be used to incorporate body movements to go along with the mix.

以下の条件で新しい"mix"を考えてください
・笑いをとれるようなユーモアのあるmix
・16小節で使用する(スタンダードmix 8回分の長さ)
・テーマは特にないが、一貫性があるとよい
・覚えやすい
・mixに合わせて体の動きを取り入れられるようなフレーズを使用してほしい

最近はやったmixの特徴として動きがある(意味不アイアイガンダムエピオン部)ものが多く、上記は覚えやすく、16小節(可変三連尺)に最近新しいのが無いと思い、僕の要求を組み込んだ。

回答

bard

Sure, here is a humorous mix that will make people laugh, uses 16 bars, has no specific theme, is easy to remember, and uses phrases that can be used to incorporate body movements to go along with the mix:

**Head, shoulders, knees, and toes, knees and toes!
Eyes, ears, nose, and mouth, open wide like a big, fat trout!
Arms up, arms down, turn around, touch the ground!
Jump up, jump down, clap your hands, shout "Hurrah!"**

This mix is easy to remember and can be used to incorporate a variety of body movements, such as clapping, stomping, and dancing. It is also a fun and upbeat mix that is sure to get people laughing and dancing.

I hope you like it!

GPT4

I’ve created a new, humorous 16-bar “mix" that incorporates body movements for a fun and engaging experience. The theme for this mix is centered around different animals and their associated movements, making it easy to remember and entertaining to perform.

  1. Monkey, Swing, Elephant, Stomp,
  2. Kangaroo, Hop, Snake, Slither,
  3. Penguin, Waddle, Gorilla, Pound,
  4. Crab, Scuttle, Bunny, Bounce!

Here’s the breakdown of the mix with the associated body movements:

  1. Monkey (make monkey gestures), Swing (swing arms),
  2. Elephant (form trunk with arm), Stomp (stomp feet),
  3. Kangaroo (hold hands in front), Hop (jump in place),
  4. Snake (move hand like a snake), Slither (body wave),
  5. Penguin (waddle side-to-side), Waddle (arms at sides),
  6. Gorilla (beat chest), Pound (pound fists on chest),
  7. Crab (sideways walk), Scuttle (move arms like crab pincers),
  8. Bunny (make bunny ears), Bounce (hop with both feet).

This mix is designed to be humorous, engaging, and memorable. Fans can shout the animal names and perform the associated movements in a synchronized manner, adding an extra layer of excitement and camaraderie to the concert experience.

考察

どちらも期待した結果は得られなかった。
GPTに関しては尺すら守られていない。

てかこんなんで笑いがとれるならお笑い芸人失業してますよ。
こういったニッチな分野でニッチな感性を必要とする分野は向こう10年ぐらいはAIに介入されない気がしてきましたね。

人間が考えるmixも飽和状態なのであえてこういったしょうもないものをやるのも悪くない気もしています。

まとめ

あきらかに趣味の領域で新しいmixを考えたところで1円も稼げない。こういったキングオブ無駄なジャンルにAIを介入させること自体が間違っていた。人間の完全敗北。

雑記