EP48｜0代碼4small time industrial business？解鎖 AI multi-generational military operation，帶你躺賺賺罐！

Then I am DaeDae, the cat who has been helping you chase AI hot searches. The AI hot search report will help you identify the key points. Do you feel that AI is developing too fast and there are new things to learn every day? Don’t be afraid. DaeDae is here to help you. Today I have selected the 5 most popular AI tool information. In 10 minutes a day, you can easily follow the AI wave and become the smartest fishing master. Before entering today’s tool introduction, DaeDae would like to talk to you about his recent mood. When I was sorting out this AI information these days, I found that I seemed to have become a kind of digital nomadic cat. In the past, we might be able to use a new software for several years, but now there are new game rules every week. I was really a little anxious at the beginning, always feeling that if I didn’t learn a certain tool quickly, would I be eliminated? But then I figured it out, AI is actually like our new paintbrush. In the past, we had to mix the paint disintegration cloth by ourselves, but now we are ready with the AI book. What we really need to practice is imagination and the ability to communicate with it. So when you listen to the program, relax and don’t be too stressed. Just treat these tools as your new toys. Okay, adjust your mood and watch the first ReadyMTAI. Many people actually don’t have any ideas. It’s that they just give up when they think of building an official website, merchandise business, shopping cart, checkout, customer information and backend management because it sounds expensive, slow, and super technical. You may just have a very simple picture in your mind, like I want to sell handmade earrings. I want to open a niche skin care product brand. I want to organize the things that my friends have been asking me to purchase on behalf of me into a formal website. But when I met ten people, I got cold feet. The interesting thing about ReadyMTAI is that it is not just an assistant to help you fill in a few lines of code, but more like an AI technology partner who will help you plan and then separate different tasks and push them forward at the same time. It can connect to Superbase, management background, order system, customer information, and will persuade these tasks to be processed together. You don’t have to complete one task before scheduling the next one. What is more practical is that it will first enter the planning mode and organize the blueprint before starting to produce content. It is less likely to complete the whole article in half. This difference is actually very big. In the past, many people were stuck not because they couldn't do it at all, but because they had to think about each step by themselves. First make the homepage, then the product page, then payment, then inventory, and finally how to organize the member information. ReadyMTAI is more like helping you draw the route first, which pages should be uploaded first, which functions should be done first, how to string together information, first have a basic stock price, and you will not keep spinning in circles. In fact, even if you don’t know how to write a program, you still have the opportunity to create the basic prototype of a boutique jewelry store in one afternoon. The homepage, store, shopping cart, checkout, order tracking, and customer management can all be up and running first. You can even use it as a test version to see if friends are willing to place orders. Is the process smooth? Is there any glitch in the screen? DaeDae’s ugly notes are very simple. The value of VerdyMTAI is to quickly organize a mess of entrepreneurial ideas into a system that can really go online. It eliminates the need for back-and-forth communication and reduces a lot of technical anxiety. Okay, stretch out and let’s look at the next tool. The second one is the update of Claude AI’s Live Artifacts. The most extraordinary thing about many AI tools is that you clearly want a small tool that can point, test, and be used directly. In the end, it only throws you a bunch of text. Or this time, Live Artifacts will directly make up for this gap. You describe the requirements in natural language, and it can make interactive small applications in the Cloud. The point is not to only make one kind. There are many types, such as habit tracking panels, morning news bulletins, team update pages, AirTable revenue trackers, social content tools, mini-games that turn PDFs into quizzes, and even e-newsletter login pages. The difference is that these are not rigid answers, but things that are really operable and can see the results, so it feels more like a mini software. It is not just AI answering you in a few paragraphs. The most impressive thing here is that it shortens the distance between writing requirements and seeing results. For example, when a content team starts work in the morning. What you most want to see is not a long list of organized reports, but a small panel with a clean screen and direct information. You can ask Cloud to help you make a morning news briefing. Put the highlights of Slack, Gmail, Google Calendar into the same page. Which meeting is coming soon, which letter is the most urgent, what things need to be processed today, and you will know as soon as you open it. Or if you are an educational content creator and have a PDF handout, you can directly ask him to turn it into a clickable and answerable gadget for students. Compared with just reading a static document, it is much more interactive. For creators, it is easier to test whether the content is really absorbed. If you are an entrepreneur, it is also very suitable to use it as a preface version. If you have an idea, you don’t necessarily have to spend a lot of money to make a complete product at the beginning. You can use it to make an operational prototype first and show it to the team, customers or investors. This is amazing. Even the lazy cat thinks of typing on the keyboard after reading it. This distance has been shortened extremely short. For people who are not good at writing programs, this is not just a convenience, but a better opportunity to really bring out the things in the brain. After learning this, I can sleep for an extra hour this afternoon. Then the third Odysseus is a free AI platform launched by the PewDiePie project. Nowadays, when many people use TrackGPT or Cloud, there will be some lags because notes, itineraries, research data, and even private files are not very secure if they are placed on other people's cloud servers. The card played by Odysseus is very straightforward. It is your AI workspace. It can run on your own computer. You can use an open source model like LAMA, or you can connect your own Cloud or OpenAI service access key. You control the whole thing yourself. The concept is very similar. You don't store all your daily necessities elsewhere. People therefore organize commonly used things in their own rooms so that they can be retrieved with ease and they know where everything is. It is not just a Yangchun chatbot. It has chats, in-depth research by agents, memories, notes, emails, routines, tasks, automatic notification chaining, and even built-in picture editing. It can also adjust the response method to make the whole environment more like a private digital work desk. But you don't have to send it as a long list of functions. You just need to capture a feeling. What it wants to do is not a single question and answer. It is an AI space where you can look forward to working for a long time. A more practical usage is that if you have sensitive customer information, internal policy notes, or you just care about privacy, you can connect the email follow-up routine without entering the original Cloud memory, do research on your own computer, build your own tool capabilities, and slowly develop an AI assistant environment of your own. For example, if you are a case consultant, the client's meeting minutes, proposal directions, and past revisions are very sensitive. You may not want these things to float around. To put it more plainly, Odysseus is more suitable for people who no longer just ask AI a few words occasionally, but who start to treat AI as a working partner. As you become more and more dependent on it, helping you organize data, assist with research, and remember context, the more you will care about where it lives, how the data flows, and whether it can be managed by itself. Although it sounds amazing, the price of Odysseus is also very honest. It is important whether your computer is powerful enough. But if you care about privacy, this is really nice because it makes AI more like living in your home. It's not about living in someone else's house. If the previous one didn't impress you, come and see this one. DaeDae, after reading it, the fourth one is not a new tool, but a super practical tip and trick, and it can be used with Gemini, QRGBT, and Cloud. Many people think that the AI ​​reply is because the model is not strong enough, but in fact, the more common truth is that the instructions you stack at the beginning are too vague and bold, and there is not enough background information. This trick is very simple, just ask the AI ​​before answering formally. If there is something unclear, ask clarification questions first, so that the interaction will not become a one-shot decision of life or death. Instead, first confirm the goals before starting to produce. When you want him to write proposals, make briefing outlines, social posts, or even create picture prompts, you can first ask him to confirm the audience, tone, format, restrictions and purpose. This small action will save a lot of repeated revisions and make the AI more like a real thinking assistant instead of a random trolling machine. You can think of it like asking someone to help with things. If you only say to organize it at the beginning, the other party can only guess first. But if the other party asks you first whether you want to organize it into a meeting version or a briefing version. Should it be short or complete? Should it be formal or relaxed? Then what comes out will usually be more accurate. The same goes for AI. Asking him to ask first is equivalent to looking at the map clearly before setting off. And this trick can be used in almost any situation. If you want to write a job search autobiography, you can first ask him to ask about your background, the position you want to apply for, and the experience you want to highlight. If you want to make a brand post, you can ask him first to confirm the target audience, whether it is students, office workers, or parents. If you want to make a presentation, you can also ask him to ask the level of the audience first, the estimated time, and whether you need to add a case. It takes an extra half minute to complete the conditions in the front, and often saves three or four rounds of changes later. This must be put into the Meatball Notes. The last fifth is the recently very topical Hermesagin Kanban-style multi-agent workflow. Many people hear the multi-agent system, and they will think of a group of AIs who work together and super smart to help you complete the task. However, the reality is often not that there are not enough agents, but that they are not coordinated with each other well. Once the process is messed up, the whole group will be stuck. The great thing about Hermesagin's approach is not to pile up more agents. Instead, we are using a Kanban board to organize work into cards, so that everything has a clear status, context, and direction. You can think of it as a work Kanban board. Each card on it represents a task. Which roots have started, which ones are being done, and which ones have been completed can be understood at a glance. This way, not only people can see clearly, but AIs also know each other better whose turn it is now and who will be handed over next. So the overall operation process looks more like an orderly operation process, not a bunch of scattered prompts. More importantly, this process can start from finding market pain points. First, different investigation agents will go to X Reddit and the Internet to find problems. Then the general control role will sort out and screen the duplicates and then hand them over to the research agent for verification. You can understand it as he helps you find the questions, verify them, drill them out, and connect them into a line, instead of having you jump in to put out the fire yourself at every step. This is where people really need help. Only the final click is for approval through Telegram. Then if you want to make a product, hand it over to the person in charge of production. Make a command line tool with the tester. If you want to make content, spit out slides and scripts. For hospital companies, small AI teams, or founders who are looking for business opportunities while making content, this is super impressive because you don’t have to keep observing the community yourself. Jumping between research, sorting, writing, development, and inspection. The real value of Hermesagin is to automate the entire AI and start to move towards a model that can operate in the long term. It’s not just for demonstration. He reminded everyone that the key is not that more agents will make you stronger, but whether the process is connected properly at each step. Let’s take a look at these five today. From VerdiMTAI to help you quickly turn your entrepreneurial ideas into operational systems, to Claude Artifacts to directly make ideas into interactive gadgets, to Odysseus, which focuses on privacy first, clarifying questioning techniques that are applicable across Gemini and ChairGPT. Finally, there is Hermesagin, which is more like a long-term operational process. The core of multi-agent working methods is actually the same, which is to make AI not only look powerful, but actually help You get things done, and you do it faster, more steadily and less tiredly. If you want to continue following this AI hot search report, remember to subscribe to the AI ​​hot search report. If today's content has gained you something, you are also welcome to help me leave a five-star review on Apple Podcast. FB IG Threads can find me by searching the AI ​​hot search report. Today's AI hot search report reports this. I am going to take a nap. We will see you tomorrow, bye meow.

那我是 DaeDae,一直幫你追AI熱搜的貓, AI熱搜報幫你劃重點 是不是覺得AI發展太快,每天都有新東西學不完?別怕 DaeDae來幫你 今天精選了5個最熱門的AI工具情報,每天10分鐘讓你輕鬆跟上 AI浪浪潮,成為最聰明的摸魚大師 在進入今天的工具介紹前, DaeDae想先跟大家聊聊最近的心情 這幾天我在整理這些AI資訊時,發現自己好像也變成了一種數位游牧貓 以前我們學一個新軟體可能可以用好幾年,現在是每周都有新的遊戲規則 一開始真的會有點焦慮,總覺得如果不趕快把某個工具學會 是不是就要被淘汰了?但後來我想通了,AI其實就像是我們的新畫筆 以前我們得自己調顏料崩化布,現在AI本我們準備好了 我們真正要練習的其實是想像力跟跟它溝通的能力 所以大家聽節目的時候放輕鬆不用壓力太大 把這些工具當成你的新玩具就好 好啦,調整好心情,先看第一個ReadyMTAI 很多人其實不是沒點子 是,一想到要做官網、商品業、購物車、結帳、客戶資料跟後台管理就直接放棄 因為聽起來又貴又慢,還超技術 你腦中可能只是有個很單純的畫面 像是我想賣手作耳環 我想開一個小眾保養品品牌 我想把朋友一直叫我代購的東西整理成一個正式網站 可是一碰到十座面 整個人就先縮了 ReadyMTAI有趣的地方在於 它不只是幫你補幾行程式碼的助手 而比較像一個AI技術合伙員會先幫你規劃 再把不同工作拆開同時往前推 它可以把向Superbase串接、管理後台、訂單系統、客戶資料 又會勸這些任務一起處理 不用一件做完再排下一件 更實用的是它會先進入規劃模式 把藍圖整理好再開始產生內容 比較不會做到一半整個跑篇 這個差別其實很大 以前很多人卡住不是因為完全不會做 而是每一步都得自己想 先做首頁 接著想商品頁 再想付款 再想庫存 最後還要凡會員資料怎麼整理 ReadyMTAI比較像 先幫你把路線畫出來 哪些頁面先上 哪些功能先做 資料怎麼串 先有個基本股價 你人就不會一直在原地打轉 實際上就算你不會寫程式 也有機會在一個下午把精品 飾品店商的基本雏形弄出來 首頁、商店、購物車、結帳、訂單追蹤 跟客戶管理都能先跑起來 甚至可以先把它當成 試水溫版本 看看朋友願不願意下單 流程順不順 畫面有沒有哪里卡卡的 DaeDae的肉丑筆記很簡單 VerdyMTAI的價值 就是把一團亂的創業想法 外速整理成真的能上線的系統 省掉來回溝通 也少很多技術焦慮 好啦,伸得懶腰 我們來看下一個工具 第二個是Claude AI的Live Artifacts更新 很多AI工具最凡的點就是 你明明想要的是一個能點能測 能直接拿來用的小工具 結果它只丟給你一堆文字 或好這次的Live Artifacts 就是直接把這個落差補起來 你用自然語言描述需求 它就能在Cloud裡面 做出可互動的小型應用 重點不是只能做一種 而是種類很多 像是習慣追蹤面板 早晨資訊意表版 團隊更新頁面 AirTable營收追蹤器 社群內容工具 小遊戲把PDF變成測驗 甚至電子報登錄頁都行 差別在於這些不是死板板的回答 而是真的可以操作 可以看結果的東西 所以感覺更像迷你軟體 不只是AI回答你幾段話 這裡最有感的地方是 它把寫需求跟看到成果之間的距離縮很短 比如一個內容團隊 早上開工時 最想看到的不是一長串整理報告 而是一個畫面乾淨 資訊直接到位的小面板 你可以叫Cloud幫你做一個早晨資訊意表版 把Slack Gmail Google Calendar的重點 放進同一個頁面 哪個會議快到了 哪封信最急 今天有哪些事情 需要先處理 一打開就知道 或者你是教育內容創作者 手上有一份PDF講義 現在就能直接請他 轉成可點擊可作答的小工具 對學生來說 比起只看一份靜態文件 互動感高很多 對於創作者來說 也比較容易拿去測試 內容有沒有真的被吸收 如果你是創業者 他也很適合拿來做展序版本 有個點子 不一定一開始就要花大錢做完整產品 可以先用他做個能操作的雏形 拿去給團隊、客戶或投資人看 呦這個厲害了 連懶貓看了都想起來敲鍵盤 他把我有一個點子 直到螢幕上真的有東西可以用 這段距離縮得超短 對不太會寫程式的人來說 這不是單純方便而已 而是比較有機會 把腦中的東西真的端出來 學會這個 今天下午又可以多睡一小時無覺了 接著第三個Odysseus 這是PewDiePie專案推出的免費AI平台 現在很多人用TrackGPT或Cloud的時候 多少都會有點卡卡的 因為筆記、行程 研究資料甚至私人文件 如果都放在別人的雲端伺服器 心理就是不太踏實 Odysseus打的牌很直接 就是你的AI工作空間 可以跑在你自己的電腦上 你可以用開源模型 像LAMA也可以接自己的Cloud 或OpenAI服務存取金鑰 整套東西由你自己掌控 這個概念很像 你不是把生活用品都寄放在別人家 於是把常用的東西好好整理在自己房間 拿取比較安心 也比較知道每樣東西在哪 它也不只是陽春聊天機器人而已 裡面有聊天 代理人深度研究 記憶、筆記 Email、行事例、排成任務 自動通知串接 甚至內建圖片編輯 還能調整回應方式 讓整個環境更像私人數位工作桌 不過你不用把它 寄成一長串功能清單 你只要抓住一個感覺就好 就是它想做的不是單點問答 而是一個你可以長期待著工作的AI空間 比較貼近生活的用法是 如果你手上有敏感客戶資料 內部策略筆記 或你就是很在意隐私 那你可以把Email跟行事例接進來 未入原本的Cloud記憶 在自己電腦上做研究 建立自己的工具能力 慢慢養成一個屬於自己的AI助手環境 比如說你是接案顧問 客戶的會議紀錄、提案方向、過往修改意見都很敏感 你可能就不想讓這些東西到處飄 再講白一點 Odysseus比較適合那種 已經不只是偶爾問AI幾句話的人 而是開始把AI當成工作夥伴的人 當你越來越依賴它 幫你整理資料、協助研究、記住上下文 你就越會在意它住在哪裡 資料怎麼流動、能不能自己管 雖然聽起來很神 但逮逮幫你抓個重點 Odysseus的代價也很老實 就是你的電腦夠不夠力很重要 不過如果你在意隐私 這個真的很香 因為它讓AI比較像住在你家 不是借住在別人家 如果剛剛那個沒打動你 來看看這個 DaeDae看完都行了 第四個不是新工具 而是一個超實用提示技巧 而且Gemini、QRGBT、Cloud都能用 很多人覺得AI回的爛是模型不夠強 但其實更常見的真相是 你一開始疊的指令太模糊太敢 背景資訊也不夠 這招很簡單 就是在正式回答前先要求 AI如果有不清楚的地方 要先問澄清問題 這樣互動就不會變成一槍定生死 而是先確認目標再開始產出 你要他寫提案、做簡報大綱、社群貼文 甚至生圖片提示詞時 都可以先請他確認受眾 語氣、格式、限制跟目的 這個小動作會少掉很多反覆改稿 讓AI比較像真的有在思考的助手 而不是隨機噴子機器 你可以想像成找人幫忙做事 如果你一開始只說包括整理一下 對方大概也只能先猜 可是如果對方先反問你 要整理成會議版還是簡報版 要短的還是完整的 要正式一點還是輕鬆一點 那後面出來的東西通常就會準很多 AI也是一樣 讓他先問 等於先把地圖看清楚再出發 而且這招幾乎什麼情境都能用 你要寫求職自傳 可以先叫他問你的背景 想投得職位、希望強調的經驗 你要做品牌貼文 可以先叫他確認目標受眾 是學生、上班族還是家長 你要做簡報 也可以叫他先問聽眾程度 預計時間、需不需要加案例 前面多花半分鐘把條件補齊 後面常常能少改三四輪 這個必須死死所進肉球筆記裡 最後第五個是最近 很有話題的Hermesagin看板式 多代理員工作流程 很多人聽到多代理員系統 腦中都會浮現一群AI分工合作 超聰明的幫你完成任務 但現實常常不是代理員不夠多 而是彼此根本沒協調好 流程一亂就整組卡住 Hermesagin這套做法厲害的地方 不是在堆更多代理員 而是在用看板把工作整理成一張一張卡片 讓每件事都有明確狀態 前後關聯跟去向 你可以把它當成一塊工作看板 上面每張卡片都代表一件代辦 哪些根開始 哪些正在做 哪些做完了 一眼就能看懂 這樣不只是人看得清楚 AI彼此之間也比較知道 現在輪到誰 接下來要交給誰 他把任務放進同一套工作架構裡 準備好的卡片 就交給對應代理員處理 能平行研究也能再收斂回來 所以整體看起來比較像 有秩序的運作流程 不是一堆散掉的提示詞 更重要的是 這套流程可以從找市場痛點開始 先由不同偵查代理員去X Reddit跟網路上挖問題 再由總控角色先整理篩選 把重複的合併再交給研究代理員去驗證 你可以把它理解成 他幫你把找題目驗證、鑽出 接成一條線 而不是每一步都要你自己跳進去救火 真正需要人出手的地方 只有最後透過Telegram按一下核准 接著如果要做產品 就交給負責製作的人員 跟測試的人員做出命令列工具 如果要做內容 就吐出投影片跟腳本 對醫院公司、小型AI團隊 或邊做內容邊找商機的創辦人來說 這超有感 因為你不用自己一直在社群觀察 研究、排序、寫稿、開發、檢查之間 跳來跳去 而Hermesagin真正有價值的地方 是讓整套AI自動化 開始向能長期運作的模式 不只是拿來示範一下而已 他提醒大家一件事 關鍵不是代理員越多越強 而是流程有沒有接好 每一步有沒有明確交棒 今天這五個看下來 從VerdiMTAI幫你把創業點子 快速變成可運作系統 到Claude Artifacts 直接把想法做成可互動小工具 再到主打隱私優先的Odysseus 跨Gemini、ChairGPT Claude都適用的澄清提問技巧 最後是Hermesagin 這種更像長期運作流程的 多代理員工作方式 核心其實都一樣 就是讓AI不只是看起來很厲害 而是真的能幫你把事情做完 而且做得更快更穩更少心累 想繼續跟著本貓摸魚追 AI記得訂閱AI熱搜報 如果今天的內容讓你有點收獲 也歡迎幫我到Apple Podcast 留個五星好評 FB IG Threads搜尋 AI熱搜報就找得到我了 今天的AI熱搜報就報導這 我要去睡午覺了 我們明天見 掰啦喵