Silicon Valley investor Zhang Lu: 70% of Gen Z's time is spent on AI applications, and traditional search has been abandoned?

Author: Lu Zhang, Silicon Valley Investor, Founding Partner of Fusion Fund

Editor: Xiaoyan Zhou

During the Chinese Lunar New Year, the explosion of DeepSeek spread to Silicon Valley, causing "anxiety" among a group of tech giants in Silicon Valley. Sam Altman, CEO of OpenAI, admitted that OpenAI's closed-source strategy has put them on the wrong side of history, and they will rethink OpenAI's open-source strategy; Zuckerberg also acknowledged in the company's all-hands Q&A that DeepSeek has achieved a novel breakthrough.

The popularity of the Chinese original large model DeepSeek also makes everyone more convinced that artificial intelligence will accelerate the change of the world.

In fact, in the past few years, artificial intelligence technology has been continuously accelerating, bringing surprises to the world.

Standing at the beginning of 2025, I am glad to share my thoughts with you. This year, I expect that five main AI trends will gradually take shape:

First, small models in vertical fields will be quickly implemented in the industry.

Second, more technological innovations at the AI infrastructure level will emerge, especially in reducing energy consumption and costs.

Third, more small models will start to enter edge devices.

Fourth, the open-source ecosystem will continue to thrive, and open-source models will be continuously enriched and take more diverse forms.

Fifth, new algorithm models and architectures will continue to emerge.

At the same time, the wide application of artificial intelligence also makes us constantly examine the relationship between AI and us. AI makes our work more "competitive" and also makes many tasks easier. With the popularization of AI tools, new competitions have intensified, but to some extent, it has also liberated our creativity, especially in the fields of art and productivity.

AI is gradually replacing some traditional jobs. But more crucially, those who can skillfully use AI tools will replace those who cannot use AI tools effectively.

Facing the rapid development of AI, we also need to have new skills and cognition. In addition to learning to use AI tools, we need to master key abilities such as asking questions and decomposing problems. The most indispensable ability in the AI era is still the decision-making ability.

Five Trend Predictions of AI

Prediction 1: Small models in vertical fields will be quickly implemented in the industry.

This is actually already happening in Silicon Valley and the industry.

In the past, everyone has been discussing large models. With the continuous expansion of the model scale, the training effect of the model is improved by accessing a large amount of data and parameters. However, in the past year, the focus has gradually shifted from the rapid growth of AI models to practical applications - that is, how AI technology can be quickly implemented and achieve large-scale industrial applications.

In this process, the large-scale application of small AI models will surely become a trend.

For many artificial intelligence applications in specific industries or vertical fields, although the amount of data needs to reach a certain scale, the quality of the data is more crucial. In fact, the quality of the data may be more important than the quantity of the data. By using high-quality industry data to train small models in vertical fields, not only can the accuracy and effect of AI applications in specific scenarios be improved, but more importantly, the scale of small models is smaller, the cost is lower, the energy efficiency is higher, and the demand for GPUs is also less.

These characteristics make small models more in line with the needs of the industrial application of AI technology at the cost level. Especially in scenarios facing enterprise customers (B-end applications), the technical cost is often the most critical factor. In this context, the "miniaturization" of the model is particularly important.

Prediction 2: More technological innovations at the AI infrastructure level will emerge, especially in reducing energy consumption and costs.

More technological innovations at the infrastructure level in the field of artificial intelligence will emerge, especially focusing on how to reduce energy consumption and costs. In fact, many related technologies have already begun to be implemented.

We know that one of the biggest challenges in the development of artificial intelligence is the high cost, high energy consumption, and excessive reliance on GPUs, which makes it unable to meet the needs of industrial-level applications. Based on this, this will drive the birth of the next wave of more practical and closer to commercial realization of AI applications.

At present, many AI infrastructure technologies can help these applications optimize the system, thereby significantly reducing the consumption of GPUs. This optimization not only includes the reduction of GPU consumption but also covers the reduction of energy consumption, fundamentally reducing the overall cost.

For example, some innovations in hardware and software technologies have been able to reduce energy consumption by 15 to 100 times or even more. At the same time, the consumption of GPUs can also be reduced to 1/4 or even close to 1/10 of the original. These are the potential trends for future development.

Prediction 3: More small models will start to enter edge devices.

In 2025, an important trend will be the application of artificial intelligence on edge and edge devices. In fact, this trend has already emerged in Silicon Valley.

The application of artificial intelligence on edge devices has been promoted by many large and small enterprises. This will give rise to new AI interfaces (interface), that is, a new form of interaction brought by artificial intelligence. In the past, our interaction methods mainly relied on traditional devices such as mobile phones, but with the popularization of AI, the future interaction methods will no longer be limited to traditional edge devices such as mobile phones. More devices will be combined with artificial intelligence, such as smart glasses, projectors, audio systems, lights, and other various daily small devices. These devices will become new carriers of artificial intelligence.

For example, a company we recently invested in - Nexa AI, its small AI model can run efficiently on edge devices such as Raspberry Pi, and its generative AI performance is similar to GPT-4. From this perspective, we can see the great potential of this technology on edge devices.

Furthermore, with the popularization of artificial intelligence, new AI interfaces will surely emerge. For example, in the C-end, AI smart glasses have attracted the attention of the industry.

In the B-end, the application forms of edge-side AI are more diverse. For example, in the logistics and supply chain industry, intelligent sensors can become the carrier of edge-side AI and be widely used in various machinery and equipment; in the booming space technology field, each satellite can serve as an intelligent agent and act as the carrier of edge-side AI; in the medical field, various intelligent medical devices, sensors, and even medical instruments can also carry artificial intelligence and become the carrier of AI.

Therefore, the carriers and interfaces of artificial intelligence are not limited to mobile phones and computers. It will be embedded in various industries, especially in the industrial field. AI will be embedded in different sensors and hardware in the form of intelligent agents. The application scenarios of artificial intelligence are very extensive.

Prediction 4: The open-source ecosystem will continue to thrive, and open-source models will be continuously enriched and take more diverse forms.

We always have a favorable view of the development of the open-source ecosystem and firmly support its growth. The open-source ecosystem has made great contributions in the field of artificial intelligence. In fact, since 2024, various open-source platforms around the world have been actively dynamic, not only in the United States, but also in China, there are many active contributors to the open-source ecosystem, such as DeepSeek, which has recently sparked heated discussions in the global tech community. The recently released Llama 3.1 also fully demonstrates that the open-source ecosystem can promote the emergence of more language models and small models.

Looking forward to 2025, I believe that the open-source ecosystem will continue to thrive and become an important platform to support and incubate start-up enterprises. Many technologies in vertical fields originated from open source, which further proves the key role of the open-source ecosystem in promoting technological progress and innovation.

Prediction 5: New algorithm models and architectures will continue to emerge.

Currently, the discussions about generative AI and large language models are very active, but at the same time, many new algorithm models and architectures are also constantly emerging. For example, the recent research papers published by Google and Microsoft are exploring new algorithm models and architectures. A significant feature of these new models is that they can not only run efficiently on GPUs, but some models can even show better performance on CPUs. This discovery may have a profound impact on the market because it raises an important question: Must all AI applications rely on GPUs? Or will there be some algorithm models that perform better on CPUs in the future?

These changes and trends are quietly taking place, indicating that they will have a profound impact on the entire industry.

AI Makes People More "Competitive" and "Relaxed"

As AI continues to penetrate into all aspects of life, the relationship between humans and AI is quietly changing. It can be said that AI makes us both "more competitive" and "more relaxed".

For a simple example, are you more "competitive" before the computer or after the computer? Obviously, it is more "competitive" after the computer. This is because various tools help us improve efficiency, liberate productivity, and thereby accelerate the progress of various tasks. With the improvement of productivity, everything is accelerating, innovation is accelerating, growth is also accelerating, and all industries are accelerating their development under the empowerment of AI. And acceleration means more intense competition. Therefore, artificial intelligence will not only promote industrial upgrading but also make us face more intense competition at the individual level.

I also want to share an interesting phenomenon. Nowadays, college students, especially freshmen and sophomores, and even some high school students, spend a lot of time using AI tools every day. About 70% to 80% of their time is spent using AI applications on mobile phones. For example, many students almost no longer use the traditional Google search but turn to platforms such as ChatGPT and You.com to search through these tools.

This change is very interesting. Because the environment we grew up in is mainly mobile-based, for us, touch screens, searches, and applications are natural and do not require special learning. For them, the application of artificial intelligence has become an instinct, similar to the process when we first used touch screens and search engines.

In addition, we find that the younger generation is more willing to use artificial intelligence tools and interact with AI. At the JPMorgan Healthcare Conference not long ago, I communicated with many heads of large companies in the medical field, and they shared an interesting phenomenon: Many companies have begun to apply AI-driven mental health tools, especially applications aimed at alleviating anxiety and maintaining mental health. These companies offer patients two options: One is to interact with AI, and AI provides feedback and suggestions; the other is to communicate online with a psychologist. The results show that 70% of people are more inclined to choose AI as the object of psychological counseling and are willing to share sensitive personal information and psychological problems with AI rather than communicating with real people. This data is surprising.

This also shows that the relationship between humans and AI may change faster than we think. From the initial lack of understanding and resistance to the gradual cooperation and now the dependence, in the future, AI may become a part of daily life like mobile phones, and everyone will form new habits.

The AI Era

What Skills and Cognitions Do Humans Need to Have

So, in the AI era, what skills and cognitions do we need to have?

First and foremost, it is crucial to realize that artificial intelligence will replace some traditional jobs, but more importantly, those who can skillfully use AI tools will replace those who cannot use AI tools effectively.

With the popularization of AI technology, new job opportunities will continue to emerge, and many traditional job positions may be replaced. In this process, the rapid development of technology will play a decisive role. Just like when we were children learning how to use a computer, today, mastering the use of AI tools has become an essential skill. In the past, knowing how to use a computer was a basic requirement for job hunting; today, mastering how to use AI tools will also become a basic ability.

(1) The ability to ask questions and decompose problems is crucial

In the AI era, having the ability to ask effective questions is particularly important. The ability to ask questions is sometimes more critical than the ability to answer questions. By asking clear and targeted questions, AI can provide us with more accurate and valuable feedback and support.

Another crucial ability is to break down complex problems into smaller tasks or to decompose a large work structure into more manageable subtasks. This is not just a skill; it reflects leadership and management capabilities. For example, in a company, one of the core responsibilities of a manager is to break down complex work into specific small tasks and then assign them to team members for execution. Now, these tasks may be assisted by AI rather than directly completed by junior engineers. If a person can only perform tasks but lacks the ability to think, ask questions, and plan, this may bring risks.

Therefore, how to think effectively, decompose problems, and collaborate with AI has become a very important ability.

(2) AI promotes the release of artistic creativity

I also observe that one of the significant advantages of artificial intelligence is its wide application in the field of art and creativity. Although some people worry that AI will replace art creators, from another perspective, AI is helping more people to create, just like the invention of the camera. The camera allows us to capture beautiful moments without the need to learn painting skills; similarly, artificial intelligence is releasing more creativity, enabling everyone to create art easily.

One of the most important abilities of an artist is to have a creative thinking and express this creativity through a strong memory and skills. Photographers did not exist before the invention of the camera, but now, mobile phones have become a tool that everyone can use to capture beautiful moments, greatly reducing the creative threshold. Artificial intelligence is developing in a similar direction, enabling more people to create and express easily.

Nowadays, even if you have not learned painting, you can present your creativity and ideas through AI. You may have rich emotions, but you may not have been able to express them in the form of paintings or songs before. Now, through AI tools, you can express your emotions in the form of lyrics, melodies, etc. This technological progress has greatly expanded the boundaries of artistic creation, allowing more people to transform their inner emotions into works of art without being limited by traditional creative techniques.

(3) Decision-making ability: The core in the AI era

The impact of artificial intelligence on various fields is obvious. Similar changes are taking place in music, art, and other industries. Taking the financial industry as an example, in the past, successful investors often relied on obtaining data that others could not obtain and making accurate market judgments through data analysis. This requires strong analytical skills, such as Excel skills and data analysis capabilities. But now, AI technology is narrowing this gap. Anyone can access more data and conduct in-depth analysis through AI.

However, the most critical and indispensable ability in the AI era is still the decision-making ability. Although AI can provide us with a large amount of information and conduct precise analysis, how to make wise decisions based on these analysis results is still the core ability. We will find that many skills, in the final analysis, are the manifestations of decision-making management ability and leadership. Only on this basis can the assistance of AI truly exert its maximum potential.

Appendix: Author Introduction

Lu Zhang, Founding Partner of Fusion Fund, a well-known Silicon Valley investor and serial successful entrepreneur, graduated from the School of Engineering at Stanford University. In 2015, she founded Fusion Fund, which