AUTHOR|PRO
E-mail|[email protected]
In today's world, when it comes to the current hot "generative artificial intelligence" (Generative AI), it is inevitable to avoid OpenAI, the hottest startup in Silicon Valley and even in the world.
But on the other hand, the characteristics of AI technology's heavy resources and high investment determine that this is not a new "sky garden", but a "pyramid" that needs to be built on the basis of the technology industry.
Founded in 1976 and 1993 respectively, Microsoft and Nvidia have been around for nearly 50 years and more than 30 years, respectively. Microsoft provides the "water source" of large-scale model cloud computing services, integrates OpenAI's technology and applications into its own products and services, and also assumes the role of OpenAI's "distributor" for global large enterprise customers, while NVIDIA is the most important creator and provider of computing power, and its constantly upgrading GPUs are continuously eaten by OpenAI's huge training network, which is transformed into stronger computing and inference capabilities and trains more powerful models.
OpenAI+微软+英伟达
The combination of "OpenAI + Microsoft + NVIDIA" has become the "iron triangle" of the industrial chain that drives the global general artificial intelligence revolution, and the young startup Open AI and the increasingly good NVIDIA have achieved each other in the platform provided by Microsoft, and together constitute one of the three most important technology companies in the world.
Microsoft, in particular, is the broader base, OpenAI and Nvidia are almost all incubated by this company, and there have been ups and downs in computer history - especially in the mobile Internet wave ruled by Steve Jobs and iPhone, Google and Android, Microsoft has almost experienced an era of "aphasia", and finally led the technological revolution again with the opportunity of the change of the personal computing paradigm and the explosion of general artificial intelligence, and its performance and stock price have been renewed.
Microsoft's story can be used as a model to examine the self-worth revolution of the world's "old computing companies", in the contemporary era when computing has become an infrastructure like water, electricity and coal, those rising stars in the past are facing the opportunity of the era of general artificial intelligence revolution, so that personal computing and their own development glow with new vitality at the same time, complete their own value transformation to "Old Money", and truly become a broader base in the AI era - naturally, will also assume the mission of AI inclusiveness.
Who can seize the new opportunity of the turning point in history? The answer belongs to Lenovo.
From Microsoft to Lenovo: Old Money's New Vision
Bill Gates, founder of Microsoft Corporation in the United States
Before the personal computer (PC) entered thousands of households, Microsoft was the "New Money" in its entirety.
"Let every family have a computer on the table" is the "vision for humanity" put forward by Bill Gates, the founder of Microsoft, in 1980. Microsoft doesn't make computers directly, but its Windows operating system dominated the desktops of the vast majority of the world's PCs for more than 20 years. With this vision, Microsoft became the first company to emerge in the wave of the personal computer revolution.
Today, Microsoft has successfully grasped the most violent wave of technology that devours advanced computing power and computing resources - the general artificial intelligence revolution, and the baton is handed down.
Founded in 1984, Lenovo is 40 years younger than Microsoft this year.
Yang Yuanqing, Chairman and CEO of Lenovo Group
If you seriously review the history of Lenovo, it is not difficult to find that 10 years is its every evolution cycle. In the words of Yang Yuanqing, chairman and CEO of Lenovo Group, Lenovo can be regarded as a new company every 10 years.
From 1984 to 1994, the first 10 years of Lenovo, Lenovo mainly acted as an agent and distributor of foreign information technology products, which was the beginning of China's large-scale digital popularization, and also laid the foundation for Lenovo's next 10 years - the PC era.
From 1994 to 2004, it was the second 10 years of Lenovo. Lenovo may be the first group of companies in China to pay attention to "private brand", and launched its own brand of personal computers in 1994. "Lenovo computer" was China's "new quality productivity" at that time, and it was the most important driving force for the popularization of PC in China. The large-scale popularization of personal computers in China has laid the foundation for China's 20-year rapid Internetization process, and also shaped the holy grail of Lenovo's future global PC market champion.
From 2004 to 2014, it was the third 10 years of Lenovo. Lenovo's theme during this period was "globalization". Through acquisitions and international integration, Lenovo became the No. 1 player in the global PC market in 2013 and has remained so to this day. Today, 10 years later, "Made in China" and "Made in China Go Overseas" have become the keywords, and Lenovo has become the most important toolbox and methodology resource library for the globalization path of Chinese enterprises.
From 2014 to 2024, it is the fourth 10 years of Lenovo, when the upward momentum of the PC market begins to enter a new round of contraction and accumulation, Lenovo begins to expand horizontally and enter the stage of diversified development. It has become a company that has formed a layout in computing resources, which is the "computing power" that is quite popular today.
Looking back on the past 40 years, Lenovo has not changed its mission of "Pratt & Whitney" to make more people use advanced technology, and every 10 years, Lenovo has the same rhyme every change: to refresh the value foundation of its foundation with technology that anchors the future. Based on this, Lenovo and Microsoft have had a lot of tacit understanding over the past 30 years.
First, Microsoft and Lenovo each played a key role in the popularization of the PC in the 90s. Because of China's economic and social development, Lenovo became an independent PC brand, much later than Gates proposed that "every family has a personal computer on their desk", but at the right time – unlike the explosion and rapid peak of smartphones, the popularity of personal computers around the world and in China has not faded to this day. In this process, Lenovo not only became the most important driver of PC popularization in China, but also became the backbone of global PC adoption in the next 10 years. It can be said that Lenovo's vision of "a computer on the desk in every home" has the best response in the world.
Second, neither Microsoft nor Lenovo are the real beneficiaries of the wave of the smartphone revolution, but they have not been silenced by their loss, but have seized the key opportunity period of computing paradigm change and become key digital infrastructure providers in the global and Chinese markets. Microsoft's 2013 full-scale transformation bet on cloud computing and cloud services not only turned its earnings around, but also created new technological imagination. In its third decade, Lenovo has gradually formed a "device-edge-cloud-network-intelligence" architecture layout, becoming a computing-driven company, truly broadening the boundaries of Lenovo's business, and Lenovo has also struggled but resolutely to become a global technology enterprise with a real technical hard core.
Third, both Microsoft and Lenovo are trying to become the first wave of the general artificial intelligence revolution at key points in time from 2023 to 2024. Microsoft is not only one of Open AI's most important shareholders, but also the largest global distributor and technology implementer of ChatGPT, and one of the key players in the new smart hardware species of AI PC.
On the eve of the new 10-year development, "general artificial intelligence" has become a natural ability after Lenovo's 40-year layout, and as Microsoft has become a key player in the AI era in the technology and application layer, Lenovo has also accelerated its development and become the fastest company to launch a new type of personal computing device - AI PC, becoming the first product to face the great prosperity of human general artificial intelligence. And in terms of definition, standard development, and product development and shaping of AI PCs, Lenovo is almost the world's fastest-moving and most determined hardware company - you can even see Lenovo launch this month, the world's first true AI PC new species.
AI PC: The endpoint is the digital twin
Lenovo Kickoff' 24
Compared with the technology companies that have shone in the era of Internet popularization, the old technology giants that have entered the industry in the personal computer era, as well as their founders who have faded out of the public eye, often have a more acute understanding and reaction to the general artificial intelligence technology revolution - this is a renaissance belonging to the computing era.
Bill Gates called artificial intelligence "the only revolutionary technological advance" he has experienced in his lifetime, and one of the most active venture capitalists in Silicon Valley's general artificial intelligence space is Vinold Khosla, one of the founders of SUN, a computer server company founded at the same time as Microsoft, and a founding partner of Khosla Ventures.
The fundamental reason for this may be that the demand for AI is so similar to the era created by this group of people: the processors that carry computing tasks have shifted from CPUs to GPUs and then to hybrid computing, just like the board combinations of the past, the computing environment has shifted from self-deployed servers to cloud computing servers, just like the germination of the Internet, and the main language for developing new scenarios and applications has changed from programming languages to natural languages.
Computers are ever-changing, but fundamentally inseparable from the combination of "display + host"; notebooks have different shapes, but the core is still the folding + highly integrated product model; smart phones are blooming, but the form created by the iPhone is still the basic solution for people's cognition and use, and the AI PC is about to come out, and it can't escape this basic law: establish a standard experience paradigm, and establish application standards from top to bottom by providing rich and large-scale products, thereby triggering a productivity blowout brought by new technologies.
Here are a few key words, the first is "rich and scalable". This is enough to keep most competitors out, AI breaks the situation that computing equipment is an island, needs the right infrastructure to carry it, and needs enough and variety of personal devices as a window of use, which means that the ability to adhere to globalization and a large and complete product system make competitors few in an instant.
In order to connect and collaborate with hundreds of millions of computing devices, enterprises will need to increase investment in hybrid infrastructure and hybrid AI, which requires true "knowhow" innovation and inclusive architecture design, as well as sufficient experience.
With this in mind, it's not hard to understand why AI PCs are a new species, and why it's not so easy to develop and launch an AI PC. This is not only a huge blue ocean market, but also a moment of accumulation before the explosion of revolutionary technology, but it needs strong and experienced players - no experience in the research and development of personal computing equipment, no strong "cloud-end" hybrid computing infrastructure development and deployment capabilities, and no strong supply chain support.
"Let everyone have an AI PC on their desk"
Let's look at this key question again: the old companies that were massively popularized in the last generation are often lost in the new era of success, but some of them are able to rely on their strategic vision and resource capabilities to find the hidden success code of history and the present, and rejuvenate and revitalize.
As for Lenovo, as China's almost the oldest and most global technology company, at the moment of the outbreak of the wave of general artificial intelligence revolution and the birth of a new species of AI PC, it will rewrite the vision of "making every family have a personal computer on the desk" put forward by Gates 44 years ago.
It can be "upgraded" to: let everyone own their own digital twin.
ThinkBook transparent screen laptop
The first step of this vision, so that everyone has an AI PC by their side, is the most direct and effective way to achieve "AI for everyone". If state-of-the-art AI has been ChatGPT and its peripheral services that open in a dialog box for $20 a month, rather than being "internalized" into a device and interface that is displayed at a person's fingertips, it will always be a long way from "AI accessible to everyone."
Based on this understanding, looking at the uniqueness of AI PCs from the perspective of "computing" makes it easier for us to understand what kind of new species AI PCs are:
AI PCs are PCs driven by large language models, and the operation of large language models on PCs is based on local heterogeneous AI computing power formed by a combination of CPUs, GPUs, NPUs, and more accelerators. As the best carrier of current AI applications and practices, the essence of the PC is a powerful personal computing device and a productivity creation tool, as well as the fundamental starting point of personal computing, and human beings are now at a new beginning of the personal computing revolution. Your AI PC is your digital twin embodied robot.
The situation is even more complicated when you cram a large language model into a PC that is dedicated to an individual and holds personal documents, data, biometric information, and usage habits—no one wants to "dedicate" a complete self to a large model that relies entirely on the cloud environment for computing and reasoning. On the device side, that is, the computing and especially inference capabilities of the PC itself, become a must. Therefore, through the technical means to ensure privacy and data security, the personal model is combined with the cloud public model through the network to achieve better computing, and at the same time, it continuously improves and feeds back the growth of the model, and then forms a personal and intelligent AI assistant.
In other words, the real mission of AI PC is to provide everyone with the "personal big model" they really need - it doesn't need Lenovo or other PC manufacturers to recreate a ChatGPT, Wenxin Yiyan or Tongyi Qianwen, but it needs a PC manufacturer to "compress" these mainstream large language models into one In the PC, make it smaller to reduce its computing and inference costs, ensure that the user's personal data and privacy are not uploaded to the cloud for inference, and also allow the device-side model to access the computing and inference capabilities of its "parent model" in the cloud environment at any time, so that "what should be implemented on the device side is implemented on the device side, and what must be completed on the cloud side is completed on the cloud side".
An AI PC is a digital twin of a person. Using the popular concept of "AI Agent", everyone can use their own AI PC as their own AI Agent, an intelligent partner. For Lenovo, its upcoming AI PC, including CPU, GPU, NPU local heterogeneous AI computing power, also on the premise of ensuring user data and privacy security barriers, has a built-in personal large model compression technology, and supports the creation of an open AI application ecosystem, under the architecture of "device-edge-cloud-network-intelligence", so that the "personal large model" can really become a person's digital twin at any time through cloud-side computing, device-side reasoning, extranet retrieval, personal database call, etc.
Allowing everyone to have an AI PC with a digital twin is the real long march behind "AI Inclusion".
Let everyone use AI, what Lenovo is doing
Lenovo Kickoff' 24
Returning to the AI technology spire of "OpenAI + Microsoft + NVIDIA" at the beginning of the article, we now clearly understand that the application under the technology is a larger and more valuable application layer, and what Lenovo needs to do now is to invest in the positive cycle of R&D and manufacturing as soon as possible, which is the first mission of technology inclusiveness.
It can be seen from Lenovo's investment in R&D, according to statistics, Lenovo Group's R&D personnel have increased to nearly 19,000 in the last fiscal year, and the R&D expense rate is expected to hit a record high, forming a "resilience" to start, a multi-engine, diversified development pattern, and the formation of organizational strength.
This is also the best echo of the will of the new era: in 2024, "artificial intelligence +" will be written into the government work report for the first time. The report also proposes to deepen the R&D and application of big data and artificial intelligence, carry out the "artificial intelligence +" action, and build a digital industrial cluster with international competitiveness. Artificial intelligence is already an important engine for the development of new quality productivity, and according to data released by the Ministry of Industry and Information Technology, in 2023, 15% of enterprises in the country have adopted generative AI, with a market size of more than 14.4 trillion yuan.
Through AI inclusiveness, especially AI PCs into thousands of households, Lenovo can really promote a new round of computing upgrades while unleashing the endless imagination and spatial value of the AI industry.
This is the best starting point for Lenovo's new 10 years, and it is also the best start of the AI era.