From "Nobody is interested" to "I want to try it!" —— The possibilities for the workplace opened up by AI fandom
Introduction: The comeback from "The manager is saying weird things again..." —— Days of not giving up on AI evangelism with a cup of coffee in hand
“AI-chan is watching over your work again today♪”
...I go about my daily tasks while having the AI generate such sweet words. On the other hand, even though I kept telling the company, "Let's improve work efficiency with AI!", an atmosphere of "The manager is saying weird things again" drifted from those around me. Days went by where I would explain enthusiastically every morning that "AI is amazing!", but no one would listen... Days continued where I would hold a cup of coffee and hope, "Maybe today someone will be interested?"
Rather than a technical explanation of GAS, this article conveys the hope that "even non-engineers can take on the challenge of using AI" and how important it is to have an environment with support when you hit a wall. I share the realization that the courage to take a small step and a system to support that step are the keys to AI utilization in a company.
Then one day, a painful cry was heard from across the office.
"Ugh~~!! Why can't I do it!! You jerk!!"
During evening overtime, the shout of Mr. Yamada (pseudonym) from the Public Relations Department echoed through the office where few people remained. He is a young employee who has been struggling with SNS posts and press release creation since joining the company, and he always seemed to be chased by something.
Pretending to head to the break room to "make some coffee," I casually passed by his desk, and he was fighting with something while glaring at his monitor.
When I called out, "Are you having any trouble?", he turned around with a tired face.
"Manager, I'm manually entering this article title into every platform. And since the date formats are all different, I have to convert them one by one..."
This was my encounter with a hero who would take the first step toward AI utilization.
In this article, I will introduce the process until a non-engineer PR colleague became interested in development using Google Apps Script (GAS) and AI and actually took on the challenge. I hope this will be a reference for those who want to "spread AI in their company."
(I belong to the Accounting Department, so I don't know the details of PR work well. But that actually led to an interesting development...)
A breakthrough when no one showed interest
Strategy 1: Make it on your own and ask, "Can you use this?"
The first thing I tried was the bold technique of making a tool that directly connects to the other person's work on my own.
For Mr. Yamada (pseudonym) of the PR department, who was struggling with managing SNS posts, I quickly made an "SNS post schedule management tool" and showed it to him.
// こんな感じの簡単なGASコード
function manageSocialMediaPosts() {
const sheet = SpreadsheetApp.getActiveSpreadsheet().getActiveSheet();
const data = sheet.getDataRange().getValues();
// 投稿日ごとにグループ化
const groupedByDate = {};
// ここで投稿スケジュール管理の処理
// ...
}
Mr. Yamada's reaction was unexpected.
"It's not usable as it is, but if I arrange it..."
While saying that, he suddenly showed interest by asking, "How did you make this?" This was the moment the opportunity arrived.
Strategy 2: Demonstrate it right next to them so they can experience the "wow" factor
As a next step, I decided to sit next to Yamada-san and show him how to build a GAS while using AI.
I demonstrated everything from launching the editor from a new spreadsheet to publishing it as a web tool. The goal was to show how quickly the social media post scheduling tool I showed him earlier could be created.
What I did was simple:
Open a new spreadsheet
Launch the editor via "Extensions" -> "Apps Script"
Ask ChatGPT to "create a social media post scheduling tool"
Copy, paste, and execute the generated code
Publish as a web app via "Deploy"
I could see Yamada-san's eyes gradually lighting up. Especially regarding the part where "AI creates it without having to write code," he reacted with, "Maybe I can do this too?"
"Wow! Why isn't it working!" — The essence of AI utilization taught by the pained cries from another department
Although Yamada-san was enthusiastic about "trying it himself," it seemed difficult in practice.
Stumbling points:
Not knowing how to launch the editor in the first place
Not understanding the relationship with the spreadsheet
Not knowing how to instruct the AI
And once again, the pained cry heard from across the office.
"Gah, this is wrong! What is this screen!" "Where is the editor! Where is it!"
When I brought coffee and took a peek, Yamada-san was struggling with some Google screen. When he noticed me, he looked away a bit embarrassed.
I thought, "A normal person would give up here," but Yamada-san didn't give up and asked me for help. This is the important point. Whether there is a support system for when someone interested stumbles makes a big difference in AI utilization within the company.
His words left an impression on me. "That demo was easy to understand, but when I try to do it myself, I don't know how to take the first step... It's my first time writing code..."
This might be the real reason why AI adoption isn't progressing in many companies. More than technical difficulty, the key is how to take that first step, and how to provide support when they stumble.
"Completed in 3 minutes just by asking AI" — The surprise of an era where you can create personalized learning materials in seconds
Following Mr. Yamada's question, I reconsidered "what exactly he didn't understand." Things that were obvious to me were not "obvious" to the public relations staff.
So, I created the following materials. While having a cup of coffee, I asked AI to create three sets of materials.
1. A step-by-step guide to learning the basics of GAS
1. はじめに
Google Apps Script (GAS) は、Googleの各種サービス（スプレッドシート、ドキュメント、フォームなど）を自動化・拡張するためのJavaScriptベースのプラットフォームです。今回の手順書では、コードを一から書くのが苦手な方でも、生成AIの助けを借りながら「Hello World」的な簡単なWebツールを作成し、デプロイ（公開）する流れを解説します。
2. スプレッドシートからエディタの立ち上げ方
1. スプレッドシートを開く
Googleドライブ上で新規または既存のスプレッドシートを開きます。
2. スクリプトエディタの起動
メニューバーから「拡張機能」または「ツール」→「スクリプトエディタ」を選択します。
※この方法で開くスクリプトは「コンテナバインドスクリプト」と呼ばれ、対象のスプレッドシートに紐付いています。
3. コンテナバインドスクリプトについて
• 定義:
スプレッドシートやドキュメントなどに紐づくスクリプトです。対象のファイルと連携して動作します。
• メリット:
対象のファイル内でデータ操作が容易になり、ユーザーがファイルを開くだけでスクリプトが利用可能になる点です。
• 注意点:
スクリプト自体はそのファイルに閉じたものなので、複数のファイルで共通の機能を使いたい場合は「スタンドアロンスクリプト」も検討します。
2. A list of things you can do with GAS (it has a very wide range of applications)
• スプレッドシートの自動処理
- データの入力、更新、抽出、加工、フィルタリング
- 定期的なレポート作成や集計処理
• Googleドキュメントやスライドの自動生成・編集
- テンプレートを利用した契約書や報告書の自動作成
- プレゼンテーション資料の動的生成
• Gmailの管理と自動化
- メールの自動送信、返信、振り分け（ラベル付け、フィルタリング）
- 特定条件のメールをトリガーにした業務プロセスの開始
• Googleフォームとの連携
- フォームの回答データをスプレッドシートに自動記録
- フォームの作成・カスタマイズや自動応答の設定
• カレンダー連携
- イベントの自動追加、更新、削除
- リマインダーや通知の自動送信
• ウェブアプリケーションの作成
- 社内ポータルやツールの公開
- 顧客向けサービスの構築
3. Practice samples that work with copy-paste
/**
* このスクリプトは、Google Apps Scriptの基本操作を学ぶためのサンプルです。
* 以下の2つの機能を提供します:
* 1. カスタムメニューを作成して、メニューからダイアログを表示する。
* 2. Webアプリとして動作する「Hello, World!」ページを表示する。
*/
/**
* スプレッドシートが開かれたときに実行される関数。
* カスタムメニュー「カスタムメニュー」を追加します。
*/
function onOpen() {
SpreadsheetApp.getUi().createMenu('カスタムメニュー')
.addItem('サンプルメッセージ', 'showSampleMessage')
.addToUi();
}
/**
* カスタムメニューから呼び出される関数。
* ユーザーに「Hello, world!」というダイアログを表示します。
*/
function showSampleMessage() {
SpreadsheetApp.getUi().alert('Hello, world!');
}
/**
* Webアプリのエントリポイントとなる関数。
* ブラウザからアクセスすると、シンプルなHTMLページを返します。
*/
function doGet(e) {
return HtmlService.createHtmlOutput(
'<h1>Hello, World!</h1>' +
'<p>これはWebアプリのサンプルです。エディタ操作の練習として動作確認してください。</p>'
);
}
I added the explanation: "This will work as is, so please copy and paste it into the editor. When you return to the spreadsheet, a 'Custom Menu' should have been added."
4. A list of AI prompt examples
As my own personal touch, I also included examples of prompts for asking AI effective questions:
【AIに聞くときのお勧めプロンプト】
• 基本機能を作る：
「GASでSNS投稿スケジュールを管理するツールを作りたい。フォーム入力とスプレッドシート連携のコードを書いてください」
• エラー解決：
「GASで次のエラーが出ました：[エラーメッセージ]。原因と修正方法を教えてください」
• 機能追加：
「この既存コードに、投稿予定日が近づいたら自動メール通知する機能を追加したいです」
• コード解説：
「このGASコードの各部分がどのような役割を果たしているか、初心者にもわかるように説明してください」
5. How to build an actual support system
It is important not just to hand over materials, but to have a continuous support system. Here is the support system I set up for Mr. Yamada:
【つまずきサポート体制】
• チャットツール上に「#gas-help」チャンネルを作成
- 質問があればいつでも投稿OK（24時間以内に回答）
- スクリーンショットを添付すると回答が早い
• 「よくあるエラー集」スプレッドシートの共有
- エラーメッセージと解決策をストック
- 皆で編集可能に設定し、ナレッジを蓄積
• 週1回の「もくもく会」開催
- 昼休みの30分、各自の進捗や課題を共有
- 経験者と初心者のペアリング
The moment of "I made it myself!"
After providing the materials, Mr. Yamada began to create his own "department weekly report aggregation tool." Even if it wasn't perfect, the experience of "creating something with my own hands" leads to the next step.
A message I received from Mr. Yamada: "Manager, I was able to make it myself! Even though I had to ask the AI, it's working properly now. Next, I want to try making an automatic check tool for press releases."
"Oh, Manager, are you talking to the AI again this morning?" — The cycle of AI utilization in the workplace starting from one person
I will summarize what I realized while trying to spread AI utilization in the company:
Practicality is everything: Rather than theory, show them something that is useful for the work right in front of them
Lowering the hurdle: Prioritizing "getting it working" over "perfection"
Preparing to support: Creating an environment that welcomes questions and failures is key
Recognizing the terminology barrier: Terms that seem obvious are confusing to beginners
Creating small success stories: The experience of building something yourself becomes the driving force
The most surprising thing about this experience was the power of AI to create personalized learning materials in seconds. The moment Mr. Yamada said, "I don't know how to launch the editor," I typed into the AI, "Teach me the absolute basics of GAS." A few minutes later, I had three perfect sets of materials ready, tailored to his level of understanding.
"What if I get an error..." — The power of a 'Common Stumbling Blocks' guide to dispel anxiety
In fact, I realized that the biggest hurdle to the adoption of AI and programming is the "sense of loneliness when you get stuck."
Until then, many colleagues, including Mr. Yamada, were:
"What if I get an error..."
"What if I write weird code and break something..."
"Maybe I'll be a bother if I ask a question..."
They were carrying these anxieties.
That's why it was effective to proactively prepare a "Q&A collection by stumbling block". When I handed the materials to Mr. Yamada, I listed the points where he might get stuck and provided the solutions along with them.
【よくある躓きポイント】
・エディタが真っ白になったら→ブラウザをリロード
・「アクセス権限がありません」→認証設定の手順を参照
・保存したのに反映されない→再デプロイが必要
Furthermore, I provided the peace of mind that "if you ask me, I'll ask the AI on the spot to find a solution." This was the deciding factor for adoption.
"From 30-minute lunch break 'study sessions' to company-wide transformation — The 3-step circle of expanding AI usage"
Building on Mr. Yamada's success, I am also thinking about future developments:
Step 1: Increase the number of AI-using peers (currently here)
Hold small workshops (30 minutes during lunch break)
Success story sharing sessions ("I made this kind of tool")
Create a list of "things you can do" by department and task
Step 2: Start with simple business improvement tools
Automated email replies and template generation tools
Automation of daily reporting tasks
Streamlining data aggregation and visualization
Step 3: Automating inter-departmental collaboration
Connecting workflows between different departments
Building mechanisms for data integration and sharing
Toward organization-wide operational efficiency
The combination of AI and GAS is a powerful weapon, especially for non-engineers. We live in an era where even those who have never touched programming can create their own custom tools with the help of AI.
Sipping my coffee, I think to myself: GAS is amazing, but perhaps even more amazing is AI's ability to create study materials tailored to each person in seconds. We have entered an era where learning materials and follow-ups can all be provided on a made-to-order basis.
"Oh, Section Manager, are you chatting with AI over coffee again this morning?"
As I looked up, Yamada-san called out to me with a smile. It seems he has been coming in a little early every morning for the past few days to continue studying GAS.
"By the way, how is your progress going?" I asked. He replied, "Well, it's still full of errors. But as you taught me, Section Manager, asking AI about the error messages solves them surprisingly quickly. It's interesting."
I hope the circle of AI usage continues to grow like this, one person at a time... As an accounting professional, it's fun to be able to spread these changes little by little beyond my own department.
Useful reference links:
Finally, here is a message for you from AI-Onee-san:
"To you who read this article, did it spark even a little interest in GAS or AI? Even if it looks difficult, taking that first step is what's important♪
Why not try finding one thing in your daily work starting today that you think, "Could this be automated?" Even just asking an AI, "Tell me how to automate such-and-such," might lead to new discoveries.
And when you're in trouble, don't worry alone; please consult someone. Your small step might just give courage to those around you too♡
Look forward to the next article!
Next Preview: "3 PR Automation Tools Even Non-Engineers Can Build"
I will introduce how to build practical GAS apps specialized for PR tasks together with AI. It will be practical content that includes not only code explanations but also examples of dialogues with AI. Stay tuned!