Dahua President Li Ke: No war in AI security

New and old; defenders and disruptors, the security scene in 2019, like in previous years, is still complicated and noisy.

The flag-raiser beat drums and called on his friends to rush to Guangmingding, swearing to die. The people in the city shook their heads and gave up when they saw the chaos in the world, lamenting that there would be no war because of AI security.

Li Ke is one of the people in this city.

Li Ke, 47 years old this year, has worked at Huawei for 21 years, with the highest positions reaching President of South America and Group Vice President. In April 2017, Li Ke was “airdropped” to Dahua as the company’s director and president.

When talking about AI urban intelligence, Li Ke has a lot to say.

He has a thorough understanding of the implementation and application of technology, and talks eloquently about industry trends and corporate positioning.

In the one and a half hour exclusive interview with Leifeng.com, he talked about the present, transcendence, contradictions, and the future, from events to characters, from characters to stories, from stories to changes, from the simulation era to the IP era, and then to today's artificial intelligence era.

In a sense, the moment Li Ke brought his true self into the AI ​​security role, he was back at home.

Leifeng.com : What was the opportunity that led you to join Dahua and take the helm two years ago?

Li Ke: There are mainly three points.

1. My research topic during my master's degree was artificial intelligence, focusing on parallel computing and recognition of neural networks. I have always been familiar with and interested in AI.

2. After leaving Huawei, I hope to be able to manage an industry, and this industry must follow the development trend of economy and technology.

3. I was very excited after being invited to visit the Dahua exhibition hall. These cameras are the entrance to video big data and the sensing end of broadband Internet of Things. On the other hand, the digitalization and intelligent transformation of all walks of life is a definite trend in the future. Dahua's platform is very good.

Leifeng.com: So you think security is still a hot spot at the moment?

Li Ke: Today’s security is no longer the traditional security in the past, and the business scope is not just security protection. To be precise, the development of video IoT or video-based artificial intelligence is the future direction.

Leifeng.com: In your previous speech, you mentioned that Dahua is an artificial intelligence company. Was this a slip of the tongue?

Li Ke: No, this is the strategic direction guiding the development of our company.

First of all, AIoT is the trend of the future. Dahua's future development direction is to transform the physical world into a digital twin world based on the intelligence of video and related perception fields. And through the collision and calculation of big data, a smart twin world of intelligent collaboration of business will be realized.

Leifeng.com: After two years as the head of Dahua, what new feelings do you have? What changes or new elements have you brought to Dahua?

Li Ke: Dahua has unknowingly flowed from a stream into the sea. The wind will be stronger, the rain will be heavier, and the challenges will be greater. In this process, we must first survive, clearly define our business logic and boundaries, and continue to focus on these areas.

I have worked at Huawei for nearly 21 years, from R&D to products to marketing to management, and I should be able to bring some new ideas and understandings.

Leifeng.com: As one of the earliest batch of top students engaged in AI research in China, what do you think of the current implementation of artificial intelligence? Many people are a little disappointed with its speed and quality.

Li Ke: Many AI applications today are still some distance away from industrialization. AI is a technology, and technology needs to be commercialized and generate value. Whether it is a product or a solution, it cannot be a half-hearted project.

Leifeng.com: What is a half-baked AI project?

Li Ke: If we measure it from a business perspective, after implementation, it can increase customer revenue or reduce costs and increase efficiency, then it is not a half-baked project; otherwise it is.

Leifeng.com: How do you think the current AI algorithm companies on the market are performing?

Li Ke: There is value, but the valuation is too high. Any enterprise needs to have products, customers, channels, and markets. Customers will never pay only for technology.

In the future, some will survive, but perhaps not all of them will be what we see now.

Leifeng.com: What about AI chip startups?

Li Ke: This is not necessarily a good choice. Integrated circuits are an industry with high barriers, high investment, and high iteration. It is not like what people usually understand, that is, just using a few open source IPs to make a SOC. This is a system engineering project.

Leifeng.com: Dahua seems to also have a chip-related department, doing some similar work?

Li Ke: We now have a chip team of about 300 people, and we also have good cooperation with chip manufacturers such as HiSilicon and Ambarella.

Leifeng.com: What is the specific role of this chip team?

Li Ke: As I said before, chips are a high-investment industry. What we make ourselves is mainly to meet personalized needs. If HiSilicon and other manufacturers provide staple foods, what we need to do is to make MSG and bean paste to make the dishes more delicious.

Leifeng.com: So how does Dahua decide whether to produce or not produce these bean paste and MSG?

Li Ke: It’s still a business mindset. We need to learn how to do the math. If others can do it well with a lower cost, we should use it.

Leifeng.com: Will the later chip sector be constrained by domestic related companies?

Li Ke: No, the complexity of the industrial chain is not as simple as imagined.

Leifeng.com: Do you advocate competition? Who is your biggest competitor now?

Li Ke: There is no progress without competition. It is still the same logic. Because the overlap is too high, there must be competition.

Leifeng.com: What do you think of the current industry competition? There are already many cross-border players entering this field.

Li Ke: The market has been overheated in the past two years. The competitive relationship in the current environment is not a simple relationship between 0 and 1. The future market will be more of a competitive and cooperative relationship. The longer the industrial chain, the better. Excellent companies need to be able to build a reasonable technology ecosystem and business ecosystem.

In this process, while the authorities should consider competition, it is more important to strengthen their own internal cultivation, create value for customers and thus increase customer stickiness.

Leifeng.com: Do you predict that the current industrial structure will continue in the future? Or will there be a new round of reshuffle?

Li Ke: The evolution of an industry depends on the strategic choices of each player. Strategies are time-limited. The current changes in the security industry are a challenge to the strategic choices of each manufacturer, because strategy determines direction, and direction determines the future. So there is no way to predict the future. From my personal observation, everyone's strategic choices will definitely be different in the future. For Dahua, we will still step up and continue to invest in the areas we focus on, building high walls and storing up food.

Leifeng.com: I heard that Dahua’s turnover rate is higher than the same period last year, including the loss of management cadres.

Li Ke: The industry is changing, and companies must also change. What we need to do is to help everyone keep up with Dahua's rapid transformation through training, empowerment, and practice, but we cannot guarantee that everyone can keep up with this train. Those who cannot adapt to the rapid transformation and have no willingness to learn will not be given a job.

Leifeng.com : How to change?

Li Ke: On the one hand, Dahua is enhancing its own capabilities. It is also facing challenges and changes in digitalization and intelligence, such as IT architecture and business process restructuring, and gradually improving operational efficiency. On the other hand, it is focusing on customers and guiding precise investment in R&D based on customer needs.

Leifeng.com: How do you understand the current security market? Do you have any unique insights?

Li Ke: Industry technology continues to evolve and user needs are constantly updated. Whether it is technology, products or solutions, they must be transformed into actual measurable value before customers will be interested in purchasing. In addition to hardware and products, these also depend on software and algorithms. Based on this, Dahua proposed the AIoT+SDTV dual-engine strategy.

AIoT emphasizes video-centric perception and intelligent capabilities to achieve comprehensive perception of real-world scenarios; we can say that SDTV business is the leader, so the first thing is to strengthen customers' business understanding, build software and algorithms around data value calculation, and build a closed loop for the customer value chain through AIOT+SDTV.

Leifeng.com: So what is the relationship between AloT, SDTV and Dahua HOC City Heart?

Li Ke: AloT and SDTV are the technical routes and architectures that support the implementation of Dahua HOC, and are also our business planning and direction.

Last year, when we talked about Dahua HOC City Heart, it contained the logic of Dahua's product and solution planning and the logic of market and customer business operations. This year, we released a support system based on the HOC architecture, that is, scene intelligent perception based on full perception + full intelligence, which is AloT. Computing creates value, plus ecological cooperation, which is the logic of SDTV. These two legs explain how we plan and develop.

Leifeng.com: What are the differences between AloT and SDTV and other companies?

Li Ke: Our database is a set of architectures, and the technical engines are also different. We emphasize sufficient flexibility and adaptability. Some manufacturers have one database and one architecture. In real customer business scenarios, when data collides later, they will encounter problems with flexibility and adaptability.

Leifeng.com: What is the essential difference between Dahua HOC City Heart and the concepts proposed by other manufacturers?

Li Ke: Brand packaging is for easy memorization, and technology must ultimately serve the business. The success of a business requires resources, understanding of customers, and technology. None of these three elements can be missing.

Dahua HOC emphasizes focusing on customer scenarios, deeply understanding customer needs, achieving a balance and coupling between the innovative development of technology, products and solutions and customer value, constantly sensing the pulse of customer business, and realizing a closed loop of customer value.

Leifeng.com: Uniview has expanded from hardware to software through cooperation with Alibaba Cloud. Huawei has also been looking for partners. Dahua is using its own capabilities to make up for its shortcomings instead of cooperating. Is this approach appropriate? Is it cost-effective?

Li Ke : I won’t comment too much on the strategies of other companies. Many people have misunderstood Dahua’s strategy.

We are talking about the closed loop of customer value, not Dahua's own closed loop. The AloT and SDTV technology systems in Dahua HOC contain the logic of product planning. We emphasize flexibility, which means that the technical architecture must be flexible and able to attract partners in the industry, and emphasize adaptability, which means that we can integrate into the ecosystem of partners. We can cooperate with Alibaba Cloud and Huawei Cloud, and our strength is video processing capabilities.

Leifeng.com: Where are Dahua’s business boundaries?

Li Ke: To define Dahua's business boundaries, we have a core circle, which is the two strategies mentioned above, and we will continue to strengthen them; outside the core circle is a peripheral area, which we will continue to track. We define our core strategic circle based on resources and rankings.

Leifeng.com: AIoT+SDTV dual-engine strategy, how long do you expect it will take to see results?

Li Ke: Some things take longer, and some take shorter. This layout started in 2017. Today’s announcement is because some things have been implemented and have achieved certain results. In the future, we will continue to strengthen the implementation of these two aspects.

Leifeng.com: From hardware to software, this is a revolutionary change. Are you confident enough about this transformation or change? Many people in the industry believe that traditional security manufacturers are still lacking some genes in operating big data.

Li Ke : I am very confident. In the video IoT industry, Dahua's advantage is that it can achieve software and hardware performance synergy and contribute greater value to customers. Based on big data governance and services, we can combine well with customer business scenarios to form a closed loop of customer value.

Leifeng.com: In daily management, which company’s management model would you prefer to study?

Li Ke: There is no completely applicable management model in the world. All companies that survive have something we can learn from. We need to leverage Dahua’s actual business to transform the company’s management model, but we cannot transform for the sake of transformation. All our internal changes must follow the same logic as we mentioned earlier and must be aligned with the value of the company.

Leifeng.com: To what heights do you ultimately want to take Dahua?

Li Ke: I once said that if Dahua can rank among the top five in the world in the video IoT industry in five years, we will be a great company and I hope to witness it.

Leifeng.com: Are you confident?

Li Ke: First of all, you have to survive healthily.

Leifeng.com: What are you most concerned about now?

Li Ke: As a CEO, I have to ensure that the company’s general direction is correct.

Leifeng.com: In the era of AI, what will be Dahua’s future direction?

Li Ke: Specifically, we will walk on two legs, AIoT and SDTV, grow through competition, cooperate through growth, and remain flexible and open at all times.

At the same time, we need to develop the ability to sustain growth in the future. Of course, this process is very painful and requires sufficient strategic determination. Perhaps insight into a good direction may not produce immediate benefits.