Recently, the application ecology of large models has continued to land, and the progress has been rapid:
First, on May 13, OpenAI's official tweet announced that it would open the networking function and 70+ third-party plug-ins to all ChatGPT-Plus users; Then, on May 24, at Microsoft's annual developer conference on Build, Copilot, an AI family bucket covering software applications and cloud services, and more than 50 new plug-ins.
By May 25, Baidu said at the Mobile Ecology Conference that it would reconstruct the mobile ecology (search, library, e-commerce, etc.) with AI native thinking, AI partners, AIBOT and other functions are in internal testing, and AI notes, AI films, AI painting, etc. are about to be closed beta.
The large-model application ecology continues to land, and the intention of fully opening up existing applications through large-model plug-ins is likely to create a new traffic portal and ecology around the large model.
However, for companies with their own advantages of traffic stickiness, content (IP) exclusivity and vertical data, the landing of the big model application ecology is not an impact, but is expected to achieve cost reduction and efficiency improvement or derive new business models based on large models.
So, what does AIGC (AI Generated Content) mean in the history of science and technology, and what changes will it bring to our future lives? In addition, is there a relatively convenient way to grasp the investment opportunities contained in the wave of artificial intelligence?
Brother Ki will talk to you today, how to worry about the layout of AIGC's investment opportunities?
AIGC Big Time: From Short-Term Concepts to Long-Term Investments
Previously, the pace of rapid rise and fall of AI-related stocks showed that funds may remain more at the level of short-term concept speculation. However, as the big model continues to land, we turn our perspective to the logical level of long-term development, and we will find:
AI+ is expected to open a new round of industrial revolution, and it is still in the first half of the introduction of new technologies.
With the large model and ChatGPT-Plugin plug-in taking the lead in defining the entrance of the AI+ era, as well as the maturity of domestic large models and the landing of application ecology, the diversified software, APP and intelligent hardware application scenarios of embedded language models (ELM/EmbededLanguageModel) are expected to gradually land, AI+ gradually enters the second half, and AI+ applications gradually move from concept to growth.
Although the performance of AI+ applications has not yet been realized, the injection of capital can promote the iteration and application diffusion of the underlying technology of AI, driving the medium and long-term improvement of the AI+ economy, and the second half of the AI+ market is expected to spread to the terminal application with the large model as the entrance.
We are currently seeing that large model applications have shown good capabilities in the fields of chat Q&A, essay writing, business analysis reports, and code writing. But this is really just the tip of the AIGC iceberg.
AI painting, AI writing, etc. belong to the branch of AIGC, with the rapid development of artificial intelligence deep learning, the new generation of models can handle a variety of content formats, including text, speech, code, images, videos, 3D models, etc.
The upper, middle and lower layers of the AIGC industrial ecosystem, Source: Tencent Research Institute
This means changes from the consumer side to the industrial side:
On the consumer side, AIGC will shape a new paradigm of digital content production and interaction, and become the content production infrastructure of the Internet in the future.
On the industrial side, synthetic data drives the future of artificial intelligence, and AIGC as a productivity tool promotes the development of the metaverse and further promotes the wider application of AI technology.
If you sum it up in one sentence, it is:
AIGC may become one of the content supply paradigms in the Web 3.0/metaverse era.
AI+ Second Half: Take the big model as the entrance
The industry now uses the "fourth technological revolution" to understand AIGC, and Jigo believes that everyone can focus on this logic:
The first technological revolution was represented by the steam engine, where mechanical production replaced hands; The second technological revolution is represented by generators, and electrical appliances become new energy sources that supplement and replace steam engines; The third technological revolution, represented by computers, has led to the formation of "human-machine control systems" and advanced automation.
And the "fourth technological revolution" is likely to be a "supermachine" with unprecedented computing power, which can assist humans intellectually to a large extent.
Therefore, looking at AIGC from the perspective of the "fourth technological revolution", you will understand why Internet giants are so eager to launch their own large models and quickly promote the landing of large model application ecology:
Looking back at the previous "entrance revolutions", the Windows system lowered the threshold for PC use and promoted the Internet boom; The emergence of the iPhone and its iOS ecology have become one of the important diversion entrances in the mobile Internet era, promoting the arrival of the mobile Internet era. The provision and access of "new entrances" can not only bring prosperity to systems and applications, but also have epoch-making significance.
The process of AI technology development is too fast, especially in the past two years, the technology curve has become steeper and steeper.
In the future, the embedded language model (ELM) is expected to become the core entrance of artificial intelligence and the Internet of Everything, providing opportunities for more reality carriers to access large models and change modern human life. The second half of AI+ is expected to interpret the process of spreading to terminal applications with large models as the entrance.
History of AI technology, sources: Deloitte Research, China Merchants Securities
In the future, the embedded language model (ELM) is expected to become the core entrance of artificial intelligence and the Internet of Everything, providing opportunities for more reality carriers to access large models and change modern human life. The second half of AI+ is expected to interpret the process of spreading to terminal applications with large models as the entrance.
In the future application field, Jigo also boldly predicts:
In the next stage, the big model will be the key to reaching artificial intelligence, and the AI+ market will be interpreted to the hardware fields such as AI+ consumer electronics, AI+robotics and many AI+ application fields with the large model as the entrance.
In fact, the essence of AIGC and big models is that AI exists as a new productivity tool.
Layout of investment opportunities in AIGC
Every revolutionary advance in the history of technology has given birth to a number of fast-growing companies and comprehensive and lasting investment opportunities.
Kigo's view is:
If there is a lack of in-depth understanding of the big model and a lack of grasp of the specific technology and specific companies of AIGC, we can focus on using index funds to layout investment opportunities in the AIGC field.
The CSI Artificial Intelligence Thematic Index (930713) was released by CSI Index on July 31, 2015, which selects representative companies among the companies that provide basic resources, technologies and application support for AI as sample stocks, reflecting the overall performance of AI-themed listed companies.
At present, there are 50 constituent stocks of the CSI Artificial Intelligence Thematic Index, and the market value weight distribution is relatively balanced.
Top 10 weighted stocks of the CSI Artificial Intelligence Thematic Index
Source: China Securities Index Co., Ltd
With the continuous and rapid landing of the large model application ecology, the market will transition from the early stage of AI concept speculation to the substantial improvement of industrial application, and we may wish to use index tools to lay out long-term investment opportunities in this field.
At present, there are 3 ETF products in the market tracking the CSI Artificial Intelligence Thematic Index.
An ETF that tracks the CSI AI Thematic Index