Welcome to an exciting new era of work

It is also a history of information dissemination, a history of communication and cooperation between people, and the evolution of labor and production tools.

If labor and production activities are included in the broad concept of office, it can also be said that half of the history of mankind is the history of office and half is the history of life.

From the perspective of office, the reason why human beings can transform nature and constantly realize various dreams and ideals lies in the fact that they can continue to develop better labor tools and production methods through professional division of labor and cooperation, so as to create a better life. As long as there is innovation, it will be a never-ending process.

In this regard, human progress depends on two things. First, the scale and order of people-to-people cooperation have expanded; The second is the evolution of production tools and production methods. As long as people are more efficient in organizing together and the tools of production are more efficient, the productive forces of society will always move forward.

When mankind entered the information age, the above laws had a new expression.

In 1993, the American futurist George Gilder proposed a formula named after Metcalfe, the pioneer of computers and the founder of 3Com: V=K×N², that is, "the value of a network is equal to the square of the number of nodes in the network, and the value of the network is proportional to the square of the number of users connected to the network". (Note: V is the network value, K is the value coefficient, and N is the number of users).

This Metcalfe's law states that the more nodes, users, etc., a network is connected, the greater the value it brings. In other words, more, better and more convenient connections between people can create greater value.

Around 2018, Microsoft CEO Satya Nadella proposed the concept of tech intensity:

Tech intensity = Tech adoption × Tech capacity

(Technical intensity = technical application × technical ability)

In the wave of digitalization sparked by the outbreak of the new crown epidemic, this formula has been enriched as follows:

Tech intensity= ( Tech adoption × Tech capacity ) ^ Trust

(Technical intensity = (technical application × technical ability) ^ trust

The purpose of this concept is to illustrate that the mastery of technological capabilities and the implementation of technological applications, based on the full trust and support of technological innovation as the acceleration index, will provide a key force for future economic development.

In a 2020 speech, Nadella explicitly linked technology density, enterprise resilience and digital transformation. "For any organization to succeed in an environment of unprecedented uncertainty, it must empower its people, foster new ways of hybrid work, engage with customers in new ways, transform business models for products and services, keep employees and customers safe, and so on, which requires a continuous cycle of data to enable digital transformation and rebuild business resilience, and that's technology density," he said. ”

Nadella argues that technological tools such as cloud computing and artificial intelligence should be in the hands of every knowledge worker, frontline worker, organization, and public sector organization in the world, so as to ensure that technology is "democratized" and empowered developers to share critical capabilities, data, and information, break down silos, cross traditional boundaries, and drive innovation to achieve "economic growth for all."

Today, the development of digital technology is inseparable from the improvement of individual office productivity and collaborative office productivity. It can be said that digital platforms, digital technologies, and digital tools have become the key forces to promote human progress, innovation, and inclusive development.

In 1937, Ronald Coase proposed in his essay "The Nature of the Enterprise" that the distinguishing feature of the enterprise is that it is a substitute for the price mechanism. Outside of the enterprise, price changes determine production, which is coordinated through a series of market transactions, but there are costs to market operation, such as information discovery, negotiation, contracting, etc. In the enterprise, it is the entrepreneur who directs production and disposes of resources, which can save some market operating costs.

However, in fact, there are also a large number of communication costs and transaction costs within the enterprise, and if you don't do it well, it will produce "big enterprise disease". Therefore, enterprises must find ways to reduce internal transaction costs and remain agile and efficient through office automation, process-oriented, flattened, collaborative tools and organizational changes.

Efficient collaboration and flexible response to digital office capabilities are the necessary conditions for building an evergreen enterprise.

DingTalk, China's enterprise digital service platform with the largest number of users, described some of today's digital office scenarios to me:

In Hangzhou, the principal of a primary school said that at the end of the semester in the past, one of the biggest tasks they had was to make awards, and even broke several printers. After using DingTalk, you don't need to print anymore, you can use the low-code tools on the platform to make a small system and generate a stamp function. With such a digital certificate, parents can forward it on Moments after getting it. A proprietress of a women's clothing brand said that the company has more than 2,000 stores across the country, and every piece of clothing on the shelves is equipped with sensors. Every day, she decides whether to add orders and which ones to add.

The person in charge of Shanghai Mitsubishi Elevator said that in actual operation, no elevator is exactly the same. In the past, each maintenance worker had to bring a hundred pages of elevator drawings when he went to repair. Now with DingTalk, the maintenance worker only needs to ask the robot on DingTalk on the site, and he can immediately see the construction drawings of the elevator, and there is no need to go to the archives to find information. It used to take an average of 2 weeks to repair an elevator, but now it has been shortened to 3 days. During the epidemic in 2022, the entire Shanghai Mitsubishi Elevator system held 17,863 meetings with "nail flash meetings", with an average of 226 meetings per day, ensuring that the organization did not stop and the business continued.

In Wucai Bay, Changji, east of the Junggar Basin in Xinjiang, the Zhundong Base of the Heavy Chemical Industry Park under the Oriental Hope Group has heavy trucks entering and exiting the gate every day. Before entering the entrance, the camera will automatically perform face recognition on the driver, take photos and upload the license plate, and compare people, vehicles, and goods with the data in the system. After entering the park, the system will guide the trucks to queue up automatically and weigh them automatically. If there are many vehicles in the queue, the driver can go to eat or rest according to the system prompts, and then return according to the corresponding time, and then automatically unload the goods after weighing. Previously, a gate required dozens of people to be on duty, but now only a few people are needed.

Ye Jun, president of DingTalk, said that he was most pleased to see the changes of the digital office platform for ordinary entrepreneurs. In the past, the first thing to do when starting a business was to find office space, whether or not anyone rented it first. Now, three or two entrepreneurs have an idea, they will first build an organization on DingTalk, and then start working.

From these ubiquitous cases, it can be seen that data-driven has become the basic label of the new production and new office era. Organization members communicate online, business data is collected, aggregated and analyzed online, and enterprises are connected with upstream and downstream partners and consumers online.

If we say that the previous generation of Made in China, many of them washed their feet and went to the fields and used factories to change China; The new generation is to use digital navigation and digital tools to change China.

A new revolution is unfolding. The waves crashed on the shore and rolled up a thousand piles of snow.

In just one or two years, thanks to the launch of ChatGPT by OpenAI and the development of artificial general intelligence (AGI), the digital office revolution is rapidly migrating to the intelligent office revolution. In Ye Jun's words: "Artificial intelligence has moved from toys to tools, entered thousands of industries, and become a productivity tool to reduce costs and increase efficiency." ”

In this round of artificial intelligence tide, the United States is at the forefront again.

Bill Gates said in his blog that the term AI refers to the creation of models that are used to solve a specific problem or provide a specific service. And ChatGPT is the most revolutionary technological advancement since 1980 (the last revolutionary sign was the graphical user interface), and the entire industry will reposition around it. Businesses will stand out by how they use it.

Bill Gates believes that as computing power becomes cheaper, GPT's ability to express ideas will increasingly resemble having a white-collar worker to help you with a variety of tasks. "You'll be able to use natural language to let this agent help you with scheduling, communication, and e-commerce, and it will work on all your devices, it will look at your latest emails, learn about the meetings you've attended, read what you've read, and read things that you don't want to worry about. This will increase your productivity on the tasks you want to do and free you from the tasks you don't want to do. "On an enterprise level, enterprise-level agents will empower employees in new ways and make them more productive.

Nadella said that the development of artificial intelligence has given us the most natural user interface — one that makes it easier for computers to understand us rather than for us to understand computers. In addition to the natural interface, another breakthrough is that we now have a new inference engine. "You've got these two combined, a new user interface, more intuitive, based on natural language, but multimodal, multi-round, multi-domain, and an inference engine, almost every software category, what's productivity, what's an operating system, what's a browser, it's going to be merged in a sense."

In the past two years, Nadella talked about the application scenarios of digital technology in his speech, including:

Farmers can operate a low-cost drone to fly over the field, collect and transmit data back, and the smart cloud and intelligent edge in the farmhouse can provide instant analysis, such as where drought or pest and disease areas are high;

Operators on the factory floor, relying on next-generation technology to discern the movement of the drill bit position to ensure precision manufacturing;

No matter where they are, doctors can use augmented reality to conduct virtual consultations, examine patients' bodies, share images, and gain insights from data instantly......

On March 16, 2023, on the new outlet of ChatGPT, Microsoft launched the artificial intelligence assistant "Microsoft 365 Copilot", which is based on a large language model and can help users write documents in Word, analyze data or reports in Excel, create presentations in PowerPoint, etc.

Just as cloud computing has changed all software categories, AI will also be transformative, Nadella said, "Copilot aims to provide users with smarter office solutions through AI technology." It is able to provide personalized suggestions and solutions based on user input and needs, helping users to complete their work more efficiently."

On the other side of the ocean, the intelligent revolution is accelerating.

On February 16, 2024, OpenAI released a king-level video generation model Sora: just enter text to generate a 60-second ultra-fine HD video. This marks the breakthrough of AI technology from graphics to video. This will greatly reduce the threshold and cost of video production, so that more people can participate in the creation of high-quality videos.

Just as Sora came out, Google also announced a new generation of models: Gemini 1.5. The Gemini 1.5 Pro, the first multimodal general-purpose model, is capable of processing more than 700,000 words of text, 30,000 lines of code, 11 hours of audio, or 1 hour of video at once. This is so because the Google model uses a "multi-expert model" (MoE) algorithm, so when responding to a request, it only runs a part of the overall model, so the response is faster and easier to process.

The surging and rapid iteration of artificial intelligence technology has formed a strong support and promotion for smart office.

The era of smart office is about to emerge, just around the corner.

In the process of digital transformation and intelligent revolution, what path has China taken and where will it go? "Super Assistant - The Way of Work in the AI Era" is based on the practice of DingTalk and the development history of the global office revolution, outlining an exciting and inspiring chapter of innovation for us.

2024 marks the 9th anniversary of DingTalk. The predecessor of DingTalk was the "Correspondence" that once tried to compete with WeChat. Although it was unsuccessful, the founding team did not stop there, but found a breakthrough in the field of enterprise communications, and in 2015 created a multi-terminal platform product that specifically solves the problem of enterprise mobile communication and collaboration.

With the development of the mobile Internet, from 2015 to 2020, DingTalk has completed the accumulation of 300 million users. At this stage, DingTalk first entered the enterprise-level service software market with functions such as free calls, DING messages, and read and unread, and then relied on office automation, audio and video conferencing, Dingpan and other functions to take root in multi-person collaborative office.

Before 2024, DingTalk has had 7 iterations of major versions, all of which are aimed at lowering the threshold for Chinese enterprises to go digital. As an observer who has been concerned about digital transformation for a long time, I have visited DingTalk many times for research and had in-depth exchanges with many senior executives of DingTalk.

In their view, the first stage of digital transformation of Chinese enterprises is the organization online, the second stage is the digitalization of business, and the third stage is industrial interconnection, that is, the realization of digital collaboration between enterprises. The 7 versions of DingTalk correspond to the evolution of these stages, and are gradually upgraded.

Why can DingTalk become a foundational, platform-level foundation for the digital transformation of Chinese enterprises?

This goes back to the history of the development of China's software industry. Chinese enterprises, especially small and medium-sized enterprises, started late in the use of software, have weak payment ability, and lack payment habits, "heavy hardware, light software", the level of informatization is generally low, and there is a lack of strong opponents in China that can compete with European and American enterprise service software giants. European and American countries have a long stage of informatization, software defines everything, and the environment is very mature.

However, in the era of mobile Internet and digital economy, China has risen its own powerful mobile Internet platform, although it was in the field of consumer Internet at the beginning, but they have established strong capabilities - instant messaging capabilities, audio and video capabilities, smooth operation capabilities under high concurrency and large traffic conditions, security capabilities, etc. When they enter the industrial Internet and promote the digital transformation of enterprises, these capabilities can be provided to the majority of enterprises at a very low cost (or even free at the beginning). The huge profits in the field of consumer Internet also make these platforms not have to seek temporary returns, but are more patient to cultivate the market.

Among these platforms, DingTalk has a deeper understanding of enterprises and commerce because it was born out of Alibaba Group and has served small, medium and micro enterprises in various industrial belts for a long time. In addition, Alibaba itself has a lot of digital operation practices, from the connection of things to the connection of people to the connection of the system, and has accumulated many methods, which also helps DingTalk to promote digital transformation from a place that is easy to get started based on mature experience in the early stage of development.

But the most important reason is that DingTalk insists on focusing on the construction of the basic platform and handing over a large number of application services to partners. DingTalk has not taken a detour, and for a period of time, it also went to do specific projects and general contracting of enterprise digital transformation, but later found that DingTalk's task is to do a good job in PaaS (Platform as a Service), do a good job in basic products such as documents, audio and video, projects, and conferences, and open these products and capabilities to the entire ecosystem as a base.

Ye Jun once told me that in fact, it is not easy to do these well, just adapting to different terminal devices and corresponding to different operating systems, the workload is huge. For example, DingTalk has recently been adapting to Hongmeng, which requires 60 professional R&D engineers to work for one to two years. And when DingTalk does a good job of the foundation and "isolates" various hardware and operating systems, it is like an operating system in itself. Through it, many SaaS (Software as a Service) providers on the industrial Internet can reach users efficiently and conveniently. This is also easy to form a co-creation and sharing service ecosystem.

In December 2022, when DingTalk released version 7.0 for inter-enterprise collaborative office, the number of users had exceeded 600 million. At this time, ChatGPT has exploded on the other side of the ocean, and it is showing an extraordinary atmosphere. (Note: As of the end of 2023, DingTalk has 700 million users).

After in-depth research, DingTalk has come to two conclusions: first, the core value of enterprise digitalization has not changed, or the organization is online, business digitalization, and industrial interconnection; The second is that AI can do all of this all over again, making the experience better, the cost lower, and the efficiency higher. At the same time, AI is not only a large model, but also a small model, a medium model, a vertical industry model, and so on.

On April 11, 2023, Ali officially released the "Tongyi Qianwen" model, and DingTalk is the first product for Ali to access this model. On April 18, DingTalk opened a new world of AI with a magic wand, the slash "/".

With the slash "/", users can write titles, write poems, write stories, polish copywriting, and generate posters in DingTalk documents. Turn on the smart meeting summary in the video conference, and you can also generate agenda review, key content, to-do lists, etc. with one click after the meeting. Automatically generate a summary of chat history in IM group chats, so that new members of the group no longer have to work hard to climb the stairs; In application development scenarios such as voting apps, hand-drawn tables, and board photos, you can automatically create a low-code app with natural conversations, and modify the content of the app through conversations.

"Use AI to do everything all over again", and this is the prelude. This slash "/" has also become a super entrance to DingTalk's intelligence. By November 2023, DingTalk announced that after the internal testing of more than 500,000 enterprises, DingTalk AI's magic wand was officially launched. 17 DingTalk products, such as DingTalk Chat, Documents, Knowledge Base, Brain Map, Flash, and Teambition, as well as 60+ DingTalk AI application scenarios, are fully open for testing. By 2024, all the core scenarios of DingTalk will have intelligent capabilities.

While redoing its own products with AI, DingTalk also opens up the intelligent base AI PaaS to the ecosystem, and uses large models to help ecological partners and enterprise customers redo their products again.

Take Aiwei Electronics, a chip design company in Shanghai, as an example. The company has more than 1,000 chips with independent intellectual property rights and has to serve nearly 1,000 customers. Each product is accompanied by a product manual, usually written in English, with dozens or even hundreds of pages containing a large number of technical terms, tables, circuit diagrams, and various parameters. Customers may send online or telephone consultation at any time in use, and the company has configured seventy or eighty technical service experts for this purpose, who spend a lot of time answering questions and confirming various details from the product manual, which is very time-consuming and labor-intensive.

Based on DingTalk's AI PaaS capability, Aiwei Electronics and DingTalk co-created Aiwei's exclusive model, and built an "IC intelligent customer service" based on the model, providing customers with instant response consulting and Q&A services 24 hours a day, 7 ×hours a day. Not only has customer satisfaction been greatly improved, but many knowledge assets have also been better precipitated.

The latest progress of DingTalk's intelligent office revolution is to co-create AI Agent with ecological partners and enterprises. The office scene is extremely scattered, but the AI assistant will always be with you, from morning to night, objectively enough to portray you, and invisibly help you make various annotations. When it understands you very well, for example, when you write a travel prompt, it will read out the data of the travel criteria from the business organization you belong to, such as the maximum travel expense limit, and give recommendations within this limit. This is a very special place for DingTalk, that is, it has scenario advantages and data advantages in the office field.

In addition, DingTalk has also begun to launch a personal version for the C-end. DingTalk believes that in the era of smart office, everyone should have a "second brain" or knowledge center. For those individuals with strong office attributes, the knowledge and tools accumulated by DingTalk on the B-side, as well as the huge resources of service partners, are more qualified to create their AI assistants.

Of course, AI assistants can't be all-encompassing on day one, and they need to be slowly cultivated and polished. But its endlessly evolving learning ability and memory are undoubtedly more suitable for office work than natural people. After the B-end and C-end advance together, DingTalk seems to be doing WeChat and enterprise WeChat at the same time. No one has ever succeeded in such a model, but it's worth trying. For example, whether it is domestic enterprise WeChat, Feishu or foreign Slack, Teams, and Salesforce, they are all pure toB, while DingTalk also has a toC field, which can add friends instead of running out and leaving. It is a unique opportunity and a complex challenge that is at the same time imaginative.

For entrepreneurs, with large models and natural language dialogue, they can "climb to the sky in one step", and it is easy to clearly understand the real-time supply and marketing of people, property, and property. If an entrepreneur wants to analyze what the average sales volume of the past three months has been and which product has the highest sales, just ask. The more you ask, the more you discover.

The current economy is challenging. But as you can see from DingTalk, the creation of enterprise organizations is still very fast, thousands of enterprises will be registered every day, and two or three people can register one, and they may not be in the same city, they are in a distributed office. Perhaps, due to the development of AI, the organizational style of enterprises is changing, closer to the "emergent type", which can be built and withdrawn at any time, merged and separated, and switched quickly. This may mean that the division of labor and services of the entire social economy is further refined, and individual decision-making efficiency will be higher and the inversion will be stronger. While the pressure is high, the responsiveness and resilience of organizations are also increasing thanks to digital technology.

There is a whimsical and logical idea that the intelligent era may bring a new trend to bypass the boundaries of apps and reach directly to data. In the future, there is no need to have so many apps, just a super app on the user side. This super app will help users solve the data connection and interoperability with other apps, and will also integrate with surrounding taxi, reservation and other software. There will be no more jumps, no interfaces, it is a smart entrance, the operation is simplified, and the underlying data is opened. Even in the future, even the three layers of SaaS, PaaS and IaaS (Infrastructure as a Service) will be bypassed, and users will directly call MaaS (Model as a Service) from the AI layer, and use a MaaS interface to add models to solve all problems. At that point, "don't find a relationship to find an (AI) assistant, the assistant will help you solve the problem" will be a common scenario.

The future is here. Let's wait and see. But in any case, this is one of the most exciting new office eras. There's no reason why we shouldn't pay attention and not participate.