This video offers deep insight into how business leaders should view and use AI in an era when AI technology is pouring out like a flood. Rather than thinking of AI merely as an automation tool, it approaches AI as a way to efficiently reallocate a company's core asset, focused attention, and to accomplish a larger mission through systems. It presents a way to integrate AI into the organization like an "employee" so the company's knowledge and systems grow together, and emphasizes an AI strategy that strengthens the capabilities of the entire organization beyond individual productivity.

1. The Real Wisdom Leaders Need in an Age of AI Information Overload 🧠

Hello! This is Youngrok Chun, CEO of Moat AI. These days there is so much AI news that many people are probably feeling exhausted. But I want to tell CEOs and organizational leaders that they do not need to chase too much AI information. Most of those stories are fine to hear a few months later. They are mostly about personal AI usage, developer tools, speculative ideas that are not yet realistic, and stock-price rises at giant companies. 📉

Over the past three years, I have given AI lectures to more than 60,000 people, and especially by listening deeply to the concerns of CEOs and leaders, I have accumulated a lot of experience with how AI should be connected to real operations. Through dozens of projects our team directly participated in, I came to understand exactly what CEOs worry about most.

"The reason I say not to listen too much to the stories in the news is not that listening is bad. Of course it is good. It is because I have seen many people become more and more confused. What really matters is our own business."

Chasing AI trends in the news often increases confusion instead. The biggest problem leaders face is not a lack of technical knowledge. It is the process of redesigning the company's mission and vision: where should we go, and what can we do in this changing era? The biggest bottleneck is the question of how large a business can be imagined and what business strategy must be decided right now.

The reason to learn and study AI is not to become a technical expert.

"It is to make our company's mission bigger, more concrete, and more executable."

Each of us has our own story and works for our own customers. We have an asset called a system, built through nights of thinking and trial and error in order to satisfy customers. This system is our own way of working, something no one can take away. A well-built system creates high return on equity and acts as a moat that current AI cannot easily penetrate. Conversely, if the system is not properly established, margins are thin and work remains hard. The AI era can actually be a good opportunity to strengthen this competitiveness. Our task is to expand the business by attaching AI as a component on top of our system and our customers. 📝

2. AI as an Organizational Asset, Not Just a Personal Tool 🤝

Most AI conversations today focus on improving personal productivity or simple task automation. It is a way of experiencing AI alone, like a one-on-one tutor, and using it only for small-scale tasks. This actually increases confusion for CEOs. It makes them wonder, "Do we have to break down all the small tasks in our company one by one, learn AI anew, and attach it to each one?" But if you do it this way, the learning cost for working-level staff becomes too high, and the effect may be lower than expected. 😞

In my experience, instead of having 10 team members each use AI separately, it is far more important to design a process where those 10 people talk with one AI at the same time and work together.

"It seems much more important to imagine and design the process where an AI with an access badge comes to work at our company, listens to our context, works with us, attaches itself to our company's assets, and becomes part of our company's system."

In the end, AI usage should not disappear when an individual uses it and then leaves. We need to connect AI closely to the company in a way that lets the company's knowledge and systems grow, in a way others cannot steal. We have to experience directly how AI can be harmoniously combined with our company before we can draw the next roadmap.

Imagine talking with a spouse or friends. Instead of each person asking a different AI and sharing answers, how convenient would it be if an AI lived together in the KakaoTalk room, using shared materials, continuously recording and sharing the conversations and decisions we make? 💬 It is like looking at the same screen instead of different screens. This is the organizational form of working with AI that many people have not yet experienced. A changed way of working like this can change not only the organization's shape, but also the mission and work methods we can offer customers.

I want to define this as the question of giving AI an access badge. AI used by an individual loses its usage record, but AI with an access badge can leave everything as a record. This record becomes good source material, or data, for later AI, accumulating like historical documents. Through this, we can create better results, turn our own trial and error into an asset, and keep compounding it. 📊

3. AI That Grows Systems Instead of Reducing People 🚀

When I meet many CEOs, they often habitually say things like, "Our strategy is ultimately growth without hiring," or "We need to reduce people and cut costs." Of course, this is also an important social topic, and it is true that the quality of people who can handle AI well inside an organization matters. But I think this focus is somewhat wrong. 🙅‍♀️

The reason executives exist is not to reduce people. One of the greatest pains executives face is often called a leadership problem: directing people well. It may be a problem of people not following well, but to put it more elegantly, leaders always feel that their ability to lead many people is slightly insufficient.

Hiring people is certainly a fixed-cost issue, but the much bigger cost is communication cost. The cost and time required to persuade someone, motivate them, deliver context, align thoughts, and align feelings increases geometrically whenever one more person is added. People often say things do not go as they wish. Even people in lower positions feel frustrated when they look at colleagues below them, so how much more would CEOs feel? Hiring leaders again to lead such teams, motivating those leaders, and educating them is not easy either. All of that energy was precious energy that could have been used for customers or to grow the business faster.

Because of this fatigue, people sometimes expect AI to solve everything like a golden key. But the real question in the AI era is not how many people we can reduce, but how much better service we can provide to customers with the same people. Going further, it is a question of how large a system the same people can operate and imagine.

"There must be an enormous way to maximize the impact each person can create. Isn't that really the essence of the concern?"

4. Reallocate the Scarcest Resource: Focused Attention 💡

The scarcest resource in the world is time in which we can truly focus. The amount of focused time we have each day is extremely limited. Sometimes an entire week passes without real focus. Because one person's focused time has limits, we have no choice but to hire people. Every important change in the world happened because someone used highly concentrated time well.

On some days, a truly great song is created, an iconic movie scene is made, or something historically significant happens. Through AI, we can now secure somewhat more of this time of focused attention, which is expensive, limited, and calorie-intensive.

We all want to use a little more of this focused time for our own lives. Owners and CEOs want to escape the pain of being tied to the business, and employees would like routine work to be handled automatically and generate money so that they can use their remaining energy on family or leisure.

That is why a company's system becomes more relaxed and profitable when it requires less time in which people must deliberately judge and keep exerting focused attention. 🤩 A person who is focusing is expensive. A well-made system is one where rules and structures are so well designed that everything flows without major trouble even when people do not think hard about each step. The fewer systems you have, the harder work becomes; the more systems you have, the easier it becomes.

But building systems also requires a lot of focused attention. Life is ultimately the process of investing attention now to reduce the attention you will need in the future, and using future attention to create more compounding return structures.

In the end, executives do two broad kinds of work.

1. Designing systems
2. Operating areas that require judgment and focused attention because they have not yet been systematized

The essence of what I want to say is not that we should use fewer people. It is that we should systematize more so that people's limited focused attention can be used in better places and leverage can increase. We can turn a larger waterwheel, gather attention for more important work, and reduce miscellaneous work, overtime, and stressful tasks. Reconstructing work processes through AI is our biggest opportunity and the essence of AI technology.

"Through systematization, we want to do more work with the same people. We want to build a more productive system. We want to move into areas of work that previously only much larger organizations could handle. We can provide more and better service to our customers. We become more competitive. We can handle a larger operating range. Ultimately, we can take on a larger mission."

The larger the system one person can handle, the more people are actually needed. If the system is good, adding people creates leverage and generates much larger returns. The idea that people are unnecessary is not the essence at all. Rather, it is an area where executives should reflect on whether the system was insufficient. It is not the employee's fault; it is always something leaders should examine from their own side. 🙏

5. Dividing Work and Finding AI Opportunities in the "Middle Basket" 🧺

Another way to describe a system is rule-based: whether clear rules exist. Company work can be broadly divided into two types.

1. Work that can be systematized with simple rules (rule-based work)
2. Work with too many possible cases, where human judgment is absolutely necessary

Over the past 20 or 30 years, most rule-based work has already been automated through IT solutions, SaaS, and similar tools so that people need to spend less focused attention. But most work that has not yet been automated is work with too many possible cases. Countless IT companies have tried automation, but ultimately areas that required human hands remained.

I call this automated work back office work. It was originally work people had to think through, but once automated, it moved into support functions or back-office departments. For example, writing letters directly to customers in the past has now become automated email or repeated text-message sending, turning into a back-office area.

So what is the front office? It is the work that has not been automated, is hard to hand off to a support department, and has very important impact. Because large consequences can arise, whether risk or opportunity, it is work that smart people must handle. If you use AI, you can see that AI can very easily automate back-office-like tasks that were hard to automate in the past. 😮 Automation that once took 40 hours can now be done in 40 minutes. In the past, if a task took 100 hours a year, spending 40 hours to automate it was often not worth the cost. But if it can now be automated in 40 minutes, we can find and automate dozens of such tasks.

But the much bigger opportunity lies in breaking down the areas that remained in the front office and finding points where work we thought would never be automated can have its burden greatly reduced in an AI-native way. This stimulates our imagination and gives us an opportunity to reshape the business itself.

In conclusion, AI usage can be divided into two kinds.

1. Using AI to automate very quickly what was already automatable, such as coding
2. Breaking down work that had never been automated before, using AI to save time and focused attention, and maximizing leverage

The part that confuses CEOs most is, "Where should we start?" Should we start with small task automation? In fact, the best cases are the tasks in the "middle basket."

We can divide work into three baskets.

• Things that can be fully automated (back office)
• Things that absolutely cannot be automated (core front office: large contracts, customer management, imagination, design, and so on)
• The middle office between them, the middle basket ⬅️ where the most AI opportunities are

We need to examine this second basket closely. Things we habitually thought only humans could do can often save an enormous amount of focused attention, 60-70%, if we combine our company's rules with AI well.

For example: 📝 "Read 100 websites, summarize them, remove irrelevant content, extract only the information important to our company and our client, summarize it, and send it to the client by email." Looking at it as one large task, AI cannot do all of this alone. But if we break it down:

1. Read and summarize 100 websites
2. Remove useless content
3. Find duplicate information

If we give AI this kind of junior-employee grunt work, it can do a large portion quickly and well. If AI reads 100 websites and reduces them to 20, we can apply our focused attention to those 20, do the final review, and add a few comments. Then we can ask AI to draft the email in the format we usually use and review it again. This lets us use the attention that used to be scattered across reading, thinking, and writing entirely on "What is good content?" and "What is important information?" 🤩

For most of the work we do, it is important to divide it into at least two parts and ask, "What part can AI do better?" If it seems AI cannot do either part, divide it again into four. If you keep splitting the work, you will definitely find parts AI can do well.

AI is very good at few-shot work, meaning it can perform extremely well with only a few examples. In the old machine-learning era, hundreds of thousands of examples were needed, but now AI can handle areas requiring inference and judgment very well with only a few samples.

"The most expensive thing in the world is focused attention."

The reason we pay so much to people who focused enough to become doctors or lawyers is that their attention has accumulated. A company as a system is the same. What matters is who concentrated more to design the system, how much focused attention is currently being invested in operating the system, and how much of the customer's focused attention that accumulated attention replaces.

Ultimately, every good and service people exchange, all money in the world, is made of a single base currency: time spent focusing. If someone can do for me what would take me 200 or 500 hours to learn, then I am buying that focused attention with money I earned through my own focus. 💰

Individuals have the task of investing their own attention for self-development when they have time or money. Business leaders must think about how to reorganize and redesign the company's focused attention, where to invest the attention being exerted now so it can become thousands or tens of thousands of hours of future attention, what game plan to deliver to customers, and how far the business can expand into new markets. We need to find a way to see the entire worldview differently around our vision and mission.

Just as individuals ask, "Why do I exist? For whom do I exist?" companies also ask, "Why should this company exist? What customer problem does it exist to solve?" AI seems able to help us reimagine every moving part in our business except the customer.

Many CEOs reach a point where growth stops because of constraints in human resources, meaning the quality, quantity, and cost of focused attention. If they can find a way to reallocate resources through AI at that moment, they may find a path to achieve the far grander ambition with which they originally started the business. ✨

6. Closing 🎬

In the next video, I will show much more concrete examples and actual teams collaborating with AI, using specific tools. I am confident you will gain ideas and insights you cannot see anywhere else. This was Youngrok Chun, CEO of Moat AI. I will see you again next time! 😃