Português
EnglishRecently, Liu Run and the entrepreneurs from "Wendao Global" embarked on a trip to the United States for an exchange program.
They not only visited several international companies in person but also attended the CES tech expo, where they felt the pulse of global innovation. During this deep dialogue across the Pacific, Liu Run equipped each member with an AI tool—DingTalk A1 Voice Recorder—to record, translate, collaborate, and reflect throughout the journey.
Below is his first-hand account from the front lines:
Today, I want to sincerely share with you a methodology I’ve personally practiced for many years, along with a special companion we brought on this trip: the DingTalk A1 Voice Recorder. It’s not just a voice recorder or a translation device; it’s more like a personal AI assistant that quietly helps us capture inspiration and connect knowledge.
Where should I start?
Let’s begin with “flight time,” a constant companion during business trips.
01 Node-Based Input
Because I travel so much, I spend a significant amount of time on airplanes every year—and this “Wendao Global” trip doubled my flight hours.
Besides necessary rest, I use these high-altitude moments to read. At 35,000 feet, there are no distractions from WeChat or phone calls, making it the perfect time to dive into books that require deep thought.
But here’s the problem: how should you read?
Many people are used to “linear reading”—like playing a tape from beginning to end. Knowledge flows into the brain in a straight line, which seems natural—but has a fatal flaw: it’s hard to retrieve. When you want to find a specific passage, you often have to scan the entire book again from the start.
That’s why I prefer “node-based input.”
What is node-based input? Simply put, it’s about no longer treating a book as a single continuous stream of knowledge, but rather as a network made up of countless knowledge nodes. My task isn’t to memorize the entire network, but to identify valuable nodes while reading, mark them instantly, and store them in my “second brain.”
After all, what truly matters isn’t what you’ve read, but what you can recall.
This feels like walking through a forest. Every time I come across a unique tree, I pull out my GPS locator, record its coordinates, take a photo of its features, add my current insights, and then store this “data packet” in my personal knowledge map. From that point on, the tree becomes a node on the map that I can call up at any time.
Later, when I’m writing, making decisions, or giving a speech, I can quickly retrieve and flexibly combine these nodes.
For example, on my flight to the United States this time, I read several books about AI trends.
Some of the ideas were quite inspiring. How could I mark them instantly? If I stopped to write them down on my phone or laptop, my reading flow would be interrupted. So I tapped the small card on the back of my phone—a long press caused it to vibrate slightly, signaling, “Recording started.” I kept reading, saying, “This book mentions the idea of restraining and spreading technology, which is very interesting. The content is…” Another long press, another vibration, signaled, “Recording saved.” The whole process barely interrupted my train of thought.
This little card is the DingTalk A1 Voice Recorder we brought on this trip. Its slim design allows it to attach magnetically to the back of your phone without adding any bulk; the Type-C port eliminates the need to carry an extra charging cable.
When I arrived at my destination, I opened the DingTalk app, and the voice notes from the flight had already synced automatically, converted to text, and neatly organized into the “Reading Notes” folder. The length of each entry also helped me quickly gauge its complexity.
Next, I set a custom prompt, asking the AI to act as a “reading assistant” and automatically generate a structured summary containing “book title, core concepts, personal reflections, and related theories.”
This is a complete “node-based input” process.
I just need to focus on deciding “which trees are worth marking”; the rest—locating, photographing, and storing—are handled by the tool.
Of course, books are static, pre-organized knowledge. Once you step into the CES exhibition hall, facing a tidal wave of fragmented information, that’s when the real test begins.
How should you handle it?
02 Digital Twin Method
The CES exhibition hall is the ultimate information flood.
The scale is massive—several pavilions combined cover more than thirty football fields; the environment is so noisy that you’d almost have to shout to communicate without amplification; the information density is so high that every few steps bring three or four new products or concepts into view.
In such an environment, the brain automatically activates a filtering mechanism—psychologists call it the “cocktail party effect”: it remembers only the most stimulating and novel snippets, ignoring the rest as “background noise.”
But for business observation, this self-protection can be dangerous.
Because many critical details are hidden precisely in the conversations and displays that get filtered out.
What to do? Willpower can’t fight the information wave, and forcing yourself to remember everything is unrealistic.
So I adopted the “digital twin method.”
This concept comes from the industrial sector and refers to creating a virtual replica in the digital world that mirrors a physical entity in the real world, with 1:1 synchronization. Applied to information processing, it means: don’t rely on your brain to “remember everything”; instead, use external tools to create a traceable “digital twin” of your perceptual experience.
The brain is like a high-performance CPU, excellent at thinking, analyzing, and creating; external tools are like solid-state drives, specialized in faithful storage.
At CES, the brain’s core task should be to observe closely, engage in deep conversations, and gain sharp insights. If you force it to take on memory duties at the same time, its performance will inevitably drop—it might miss the next three important sentences just to remember one.
So as soon as I entered the exhibition hall, I removed the DingTalk A1 Voice Recorder from my phone and turned it on, slipping it into my shirt pocket.
After that, I didn’t have to worry about it anymore. Like a silent data sponge, it uses six microphones and up to 45 hours of battery life to continuously absorb all surrounding sounds: my explanations, questions from the entrepreneurs, and conversations with exhibitors.
And I focused on immersing myself in the interactions. Knowing that everything was being recorded in full actually allowed me to relax and concentrate even more.
In the end, a complete audio file titled “CES: Recordings and Explanations from the Tech Expo Visit” was created. After automatic transcription by AI, a tens-of-thousands-word transcript with timestamps and speaker labels was generated—this is my “digital twin” from CES and the raw material for my subsequent thinking and writing. Some of the content in the [article](https://mp.weixin.qq.com/s?__biz=MjM5NjM5MjQ4MQ==&mid=2651780487&idx=1&sn=e082c6571c8aaf3dead98d6fe3367c89&scene=21#wechat_redirect) from a few days ago came directly from this data set.
For example, the article mentions:
“On site, I saw a company specializing in ‘violence detection.’ Their cameras don’t recognize identities or judge expressions; they focus solely on detecting violent behavior. As soon as someone draws a knife or throws a punch, the system immediately issues an alert and, in emergencies, can even call the police directly. In China’s urban areas, this may have limited use, but in regions with weaker public safety, it’s a lifesaver.”
It’s true: everyone is talking about AI these days.
But without high-quality process data, even the smartest brains or AI models can only make low-level inferences over and over again.
First, record everything completely; then think deeply. That may be the essential skill of this era.
Creating a digital copy solves the data storage problem. But in an international setting like CES, another common challenge arises—the language barrier.
How can you break through it?
03 High-Fidelity Communication
CES offers a rare opportunity to interact face-to-face with entrepreneurs and tech experts from around the world.
But these professional conversations are extremely demanding in terms of “accuracy.”
For casual chitchat, understanding the general meaning is enough; but when discussing technical details or business models, we aim for 100% error-free communication. Because 99% understanding can lead to serious misjudgments due to just a 1% deviation.
At the exhibition this time, we encountered exactly such a situation.
At one booth, we discussed autonomous driving technology with experts from Google Waymo. Although I consider my English adequate, in high-density, highly technical conversations, I still needed to ensure that when terms like “lidar point cloud density,” “purely visual end-to-end algorithms,” and “extreme scenario testing” were mentioned, the information received was completely intact.
So I took out my phone and activated DingTalk’s “Face-to-Face Translation” feature.
I placed the phone flat between the two of us, and the screen automatically split in half: the side facing me showed Chinese, while the other side showed English, with the text automatically rotated 180 degrees for easy reading. No back-and-forth exchanges were needed—natural and polite.
I spoke in Chinese, and he immediately saw an accurate English translation; he spoke in English, and I could simultaneously read a fluent Chinese version.
Both sides were able to focus on the content itself. And the entire conversation was fully recorded by the DingTalk A1 Voice Recorder in my pocket, generating a bilingual meeting transcript that was added to the “Wendao Global” knowledge base for future reference and sharing.
You see, true technology doesn’t amaze you with its existence; it makes you barely notice it working.
Well, a full day’s worth of information has been stored in the data warehouse with “high fidelity.” But here’s the problem—the data in the warehouse is too vast and complex.
For example, when I returned to the hotel in the evening, I suddenly couldn’t remember who said a key sentence during the day or in what context it was said.
How do I find it?
04 Conversational Search
A long time ago, humans began recording their thoughts in writing to fight against forgetting. This “note-taking” process laid the foundation for civilization.
Later, notes moved into computers, and search functions greatly improved efficiency. In theory, as long as you remember the keywords, you can find any record.
But now, a more intuitive “conversational” approach to information processing is quietly emerging.
Traditional keyword searches are like looking for a book in a vast library using an index card—you have to know the title, author, or classification number. If you only vaguely remember, “Yesterday afternoon I think I saw a book with a blue cover about robotic arms,” the indexing system probably won’t help.
Conversational search, on the other hand, is like directly asking the librarian, “I’m looking for a book about the flexibility of robotic arms, and the cover seems to be blue.” With understanding and experience, the librarian can quickly help you pinpoint the target.
In other words, to enable flexible retrieval, you must first build a knowledge base that “understands human language.”
For example, one night at the hotel, I vaguely remembered a discussion about “robotic dexterous hands” that took place in the afternoon during a visit to the robotics exhibit, which I wanted to use in an article.
So I opened DingTalk and asked the AI via the chat box, “Who mentioned the view on robotic dexterous hands today?”
Within seconds, the AI located that passage from the “CES: Recordings and Explanations from the Tech Expo Visit” and summarized the key points: “Robotic dexterous hands are a key area of攻关, with the core difficulty lying in achieving sufficient flexibility and pressure-sensing capabilities…” It could even go further and answer, “What was the context of the discussion at the time?” providing a coherent narrative based on the surrounding context.
This is the power of “conversational” information processing.
It requires tools to adapt to human thinking, rather than forcing people to adapt to the logic of the tools.
I used the same approach to capture a flash of inspiration that struck me at night.
One late night, as I was getting ready to sleep, I suddenly thought, “Many of the AI startups I saw today seem to fit into the ‘perception-decision-action’ framework.”
But I really didn’t want to turn on the light and reach for my phone to record it. As I get older, sleep becomes precious, and even a small disturbance can make it hard to fall back asleep.
So I reached out and touched the DingTalk A1 button on the back of my phone. A soft buzz told me it had started. With my eyes closed, I whispered, “AI startups, perception, decision, action.” Then I fell asleep peacefully.
The next morning, this thought had already been converted into a voice note and quietly sat in the “Inspiration” folder.
Well, reading, recording, organizing, retrieving—all these preparatory steps are for the final stage—producing clear, powerful articles.
05 Scaffolded Writing Method
I believe many writers have had similar experiences: when faced with a blank document, their minds are buzzing with ideas, but they don’t know where to start.
The reason may be “nothing to say,” but it’s often because “there are too many ideas, and I don’t know how to organize them.”
And AI is particularly good at solving the latter.
Why? Because much of writing is essentially a “scaffolded writing method.”
It’s like building a house: your mind is filled with inspiration, cases, and data—these are the bricks on the construction site. But what you ultimately deliver is a structurally sound building with a unique style. From materials to finished product, the most labor-intensive part is building the scaffolding—the structure and logical framework of the article.
In the past, you had to build this scaffolding yourself, hammering out every detail in your head, which was exhausting. Now, AI can help shoulder this physical workload.
Of course, AI cannot and should not replace me as the chief architect. The soul of the article—unique perspectives, deep insights, emotional warmth, and narrative appeal—still needs to be infused by me. But AI can significantly lighten the early burden.
For example, before I start writing, I give all my “digital copies” to the AI and issue a command through DingTalk: “Please act as a senior business analyst, comprehensively analyze today’s CES interview audio files, distill 20 key insights, and generate a detailed article outline based on the framework of ‘technology trends—business applications—future outlook.’ For each insight, please include 2–3 specific cases or data points to support it.”
Soon, the AI delivered a preliminary scaffold. While it wasn’t perfect—some points still needed adjustment—it effectively solved the bottleneck of going from “zero to one.”
Next, I used this as a foundation to keep refining, gradually polishing a satisfactory outline, then filling it with evidence, fine-tuning the wording, and infusing it with emotion…
Yes, AI can’t replace thinking.
But it can dramatically shorten the distance between “thinking” and “expression.”
It lets me focus more on “creation” and less on “organization.”
Final Thoughts
From node-based input and the digital twin method to high-fidelity communication, conversational search, and the scaffolded writing method, these five steps formed my work rhythm during the “Wendao Global” trip.
You might see it as an efficient learning methodology.
The key is to free the brain from tedious “memory” and “archiving” tasks and let it focus on higher-value activities like “thinking” and “creating.”
The DingTalk A1 Voice Recorder is the physical carrier that supports this methodology. Today, I’d like to share it with you as a practical case, hoping it will offer some inspiration.
Holding the DingTalk A1 in my hand and recalling the scenes from CES, I’m gradually realizing why so many people predict that 2026 may well usher in a boom period for AI hardware.
Because the capabilities of large AI models are like electricity in a power grid—ubiquitous and full of potential. But they can only be truly utilized through various “appliances.”
AI hardware is the “appliance” that brings AI capabilities into everyday life.
At CES, AI glasses that provide instant translation, smart rings that monitor health, and AI companion devices that learn pet behaviors are no longer science fiction. Sensor, chip, and battery technologies are becoming increasingly mature, and this year’s CES theme has shifted from “software AI” to “physical AI.”
And AI hardware like the DingTalk A1 Voice Recorder is just one wave in this trend. It’s a product—and a trend—symbolizing that AI is reaching out and stepping into real life.
In the future, these tools will help professionals in various fields break free from repetitive tasks and devote themselves to more creative endeavors.
But remember: the limits of a tool are ultimately determined by the user’s imagination.
My imagination is limited. That’s why we specifically secured five brand-new DingTalk A1 Voice Recorders from DingTalk. I’m genuinely curious: if you had one, in what scenarios would you use it, and what problems would it solve? Feel free to share your thoughts in the comments. I also look forward to your engagement through likes, shares, and saves.
As of 6:00 p.m. on January 19, we’ll select the five commenters with the most likes and award them these five AI assistants.
Well, my friend,
The new year has already begun.
May you continue to learn efficiently and evolve with high energy this year.
Keep it up.
Perspective / Liu Run Lead Writer / Er Man Editor / Ge Ping Layout / Huang Jing
DomTech is DingTalk's official designated service provider in Macau, specializing in providing DingTalk services to a wide range of customers. If you'd like to learn more about DingTalk platform applications, you can contact our online customer service directly, or reach us by phone at +852 95970612 or by email at cs@dingtalk-macau.com. We have an excellent development and operations team with rich market service experience, and can provide you with professional DingTalk solutions and services!