Written on the occasion of the release of GPT4: Open the imagination space, how can Web3 accelerate the productivity revolution?

星球君的朋友们

Odaily资深作者

2023-03-15 05:13

This article is about 5615 words, reading the full article takes about 9 minutes

Explore the infinite possibilities of the combination of Web3 and AI

first level title

Core Points, TL;DR

1. AIGC will be the most important enabling technology to improve overall social productivity in the 21st century. The essence of Web3 is to reform the production relationship and return the ownership of data assets to users, which is closely related to the AI revolution.

In the Web2 era, a large amount of data is in the hands of Internet giants, and users can only become consumers of data, and cannot grasp the ownership of data. In the Web3 era, blockchain technology makes data an encrypted, secure, and traceable asset, and the ownership of the data returns to the user.

This change is very important for the AI revolution, because data is the core driving force of AI, and AI models require a large amount of data for training and optimization. Web3 allows users to master their own data ownership, better control their own data, and provide more diverse and comprehensive data for AI models, thereby accelerating the development of AI.

2. A wave of AI gadgets is emerging, but they are facing many pain points, and Web3 provides a very good solution.

AI gadgets refer to small, specialized AI applications, such as text analysis tools and image recognition tools. As OpenAI opened the API interface of ChatGPT, a large number of AI gadgets began to emerge. However, AI gadgets are still facing many pain points, including technical problems, as well as difficulties in the market and even capitalization. Web3 can provide a solution to these problems.

3. From the perspective of investment, the current focus is on Web3 as a service protocol to empower and accelerate the new native tools brought about by the productivity revolution.

At present, many applications or narratives on the market still remain: 1. Use AI to generate NFT images with one click, and then continuously enrich the scene construction of the Metaverse; 2. X to Earn mode, such as providing chatgpt with error correction and feedback, to achieve For the training of the data, the corresponding tokens are obtained. These applications are still in their infancy, but with the continuous development and popularization of technology, there are still many application scenarios that can be explored and applied in the future.

We believe that at the current stage, investing in Web3 service protocols is better than investing in specific toC tool products, especially service protocols for new tools and new revolutions, rather than inefficient vampire attacks on Web2 social networks.

first level title

1 Introduction

With the continuous development of human society, various productivity revolutions continue to emerge, which promote the progress and development of society. In recent years, the rapid development of artificial intelligence technology has brought about an unprecedented revolution, which is a productivity revolution covering almost all fields of human society. The wide application of AI technology not only greatly improves production efficiency and quality, but also brings various innovative applications and business models. However, the AI revolution is also facing many problems and challenges. One of the most important issues is how to quickly promote and implement these AI technologies so that they can become part of productivity and bring tangible benefits.

To this end, we need to find a new carrier to promote the rapid application and innovation of AI technology, and Web3 technology is a very good choice. Web3 technology not only provides a safer and more autonomous network environment, but also enables more open and decentralized applications, provides a wider range of scenarios and business models, and enables AI technology to better serve all aspects of human society.

first level title

2. Past Productivity Revolutions and Today

Past productivity revolutions can be traced back to the Industrial Revolution in the 18th century, which was a major change in human history. Prior to this, the production methods of human society were mainly manual labor and agricultural production, and the production efficiency was very low. The Industrial Revolution greatly improved production efficiency and production capacity through mechanized production, mass production, and factory organization. This transformation has completely changed the face of human society and laid the foundation for modern industrialized society.

After the industrial revolution, there have been a series of productivity revolutions such as the electric revolution and the information revolution. The electric revolution is mainly through the extensive application of electric power and the development of electrified equipment, which further mechanizes and automates the production method and further improves the production efficiency. The information revolution is mainly through the application of computer technology, which promotes the popularization of information production, Internet and mobile communication technology, making the acquisition and transmission of information more convenient and faster.

These productivity revolutions not only promoted the progress and development of human society, but also brought a huge impetus to economic development and social change. Every productivity revolution is a historic progress that brings new opportunities and challenges.

So, how will the AI revolution, and the possible productivity revolution in the future, be different from the past?

first level title

3. Web3 and AI: The combination of two revolutionary forces

Web3, also known as the Distributed Web, uses blockchain and decentralized technologies to distribute data and applications across multiple nodes, enabling a more secure, transparent and decentralized network. The mission of Web3 is to bring about a more fair and open Internet, giving users more data control and asset value.

Gus, a researcher at the Blockchain Association of Tsinghua University, believes that as a subversive innovation, AI has liberated productivity, and Web3 has reformed production relations, allowing the ownership of data assets to return to users, thereby changing production relations, so the combination of the two is promising. .

From the perspective of top-level design, first of all, AI can help Web3 to better achieve decentralization and autonomy. For example, AI can be used to automate the execution and decision-making of contracts and DAOs, improving the efficiency and reliability of Web3. At the same time, Web3 technology can provide AI with a more secure, transparent and decentralized infrastructure, making AI learning and decision-making more fair and credible.

Secondly, the combination of Web3 and AI can bring more intelligent and personalized applications and services to users. For example, a decentralized identity management system based on Web3 can provide AI with more accurate and secure data, and provide users with more convenient and secure identity authentication services. At the same time, AI can provide users with personalized recommendations and services based on their preferences and behavioral habits.

Finally, regarding the specific combination scenarios of Web3 and AI, the voices in the market generally focus on using AI's generative capabilities to empower metaverse scenarios. For example, use an AI tool to generate small pictures (connected to midjourney or stable diffusion, etc. behind it) to generate NFTs such as pictures, text, video, and task images with one click, and then place them in games and metaverse scenes for customized Production. Or, use AI to automatically generate questions to prevent cheating robots/sybil attacks. For example, readon uses chatgpt to automatically generate small quiz to prevent users from swiping coins.

first level title

4. From a market perspective, the pain points of AI gadgets

We have seen that in the recent period (early March 2023), after OpenAI opened the latest ChatGPT interface, a large number of AI gadgets have emerged on the market. According to incomplete statistics by Gus, a researcher at the Tsinghua Blockchain Association, within a week after OpenAI opened the latest ChatGPT interface, more than 1,200 AI gadgets emerged on the market, covering various workflows, including but not limited to: programming-related training model, code development; word processing tools, such as translation assistance, PDF summary (chatpdt), automatic generation of excel (chatexcel); in addition, there are many interesting gadgets, such as Xiaohongshu Copywriting Generator, Weekly Daily Generator, etc. .

As OpenAI officially opened GPT 4 in mid-March 2023, further improving production efficiency (see picture), we expect this wave of AI gadget startups to continue to emerge, and burst out multi-modal and cross-platform characteristics of the field. Multimodal means that AI gadgets can process multiple data types at the same time, such as text, images, sounds, etc., so as to achieve more comprehensive and accurate analysis and application. Cross-domain means that AI gadgets can be applied in multiple fields, such as medical care, finance, education, etc., so as to achieve a wider range of application scenarios.

There is no doubt that based on the ready-made LLM model, these AI gadgets can help us improve work efficiency, reduce costs, optimize decision-making, and even improve our quality of life.

However, there are many challenges in the development of these gadgets, and here we mainly focus on the market perspective. The business models of these gadgets are often relatively simple, mainly relying on subscription or sales to make profits. Regarding the business model of AI gadgets, the survey results show that the mainstream business model of these gadgets is to charge a subscription fee of $5 per month (such as ChatPDF). The advantage of this model is that the price is relatively low, easy for users to accept, and it can also bring stable income to developers. However, this model also has some problems.

First of all, the model of charging subscription fees has become very common in market competition, which has led to fierce competition in the market. Developers need to continuously optimize prices and functions to attract more users. This will also cause some small tools to only provide basic functions, unable to provide more advanced services, and it is difficult to highlight their own competitive advantages.

Second, the marketing of AI gadgets is also facing challenges. In terms of marketing, developers rely too much on personal influence on social media and lack real marketing strategies, resulting in higher and higher customer acquisition costs. In the case of such fierce market competition, relying solely on personal influence is far from enough to attract more users.

Finally, AI gadgets are also under-narrative, lacking a compelling story and an attractive brand image. This makes it difficult for users to establish a sense of identity with the gadget, and also makes the gadget's brand effect in the market poor, making it difficult to expand its market share.

first level title

5. How Web3 will accelerate the productivity revolution

From a market perspective, for the AIGC revolution, using Web3 can accelerate the GTM, cold start, and even rapid capitalization of these gadgets and small products. Developers can help small companies and individual developers to raise funds quickly through decentralized exchanges and crowdfunding platforms, which lowers the financing threshold and accelerates product launches.

In addition, users can enjoy most of their value in the productivity revolution-data assets. Web3 adopts a decentralized model, and realizes the decentralized storage and ownership of data assets through blockchain technology. This enables users to own their own data, rather than having it in the hands of big tech companies as it was in the Web2 era.

At the same time, DAO and snapshot voting can be used. Users can put forward their own needs, or vote for their favorite gadgets to quantify their needs. These mechanisms enable users to participate more actively in product development and improvement, improving user satisfaction and stickiness.

From a technical point of view, the decentralized and distributed nature of Web3 provides better support for AI gadgets. In the Web3 ecosystem, all data and applications are stored on the decentralized blockchain network, and these data are open, transparent, and non-tamperable. This data architecture makes it easier for AI gadgets to acquire, share and manage data while ensuring data security and privacy. In addition, the smart contract function of Web3 can also provide more flexible and efficient transaction, cooperation and governance mechanisms for AI gadgets, making the application and management of AI gadgets more convenient and effective.

These features help improve the performance and reliability of AI gadgets. Using distributed computing, storage and communication technologies, Web3 can provide AI gadgets with a more efficient, reliable and secure computing and communication environment. For example, Web3's IPFS distributed file system can provide AI gadgets with faster, reliable, and decentralized data storage and sharing services, reducing data delay and loss. The P2P communication protocol of Web3 can also provide AI gadgets with faster, reliable, and decentralized communication services, and support real-time interaction and collaboration between AI gadgets.

Again, this is a rare opportunity for Web3.

first level title

6. Typical applications of Web3 accelerating productivity revolution

As mentioned earlier, we found that many applications or narratives currently on the market are still stuck in:

1. Use AI to generate image NFT with one click, and then continuously enrich the scene construction of the Metaverse;

2. X to Earn mode, such as providing chatgpt with error correction and feedback to achieve data training and obtain corresponding tokens.

These applications are still in their infancy, but with the continuous development and popularization of technology, there are still many application scenarios that can be explored and applied in the future.

Foresight Ventures' article "Web3 Application Layer Innovation under the AIGC Wave" mentioned an example of using the Web3 service protocol to try to achieve this goal. Here we briefly introduce the product of the Tsinghua University and MIT teams, Tie Protocol. As a Web3 service protocol for disruptive innovation technologies such as AI, Tie Protocol is positioned as a Web3 protocol for disruptive innovation technologies. Its goal is to explore new Web3 native application scenarios, rather than just occupying the scenarios already occupied by Web2. Thus providing higher ROI and value. To some extent, the positioning of Tie Protocol can be regarded as "Apple Store or WeChat Mini Program Mall for Web3 native applications".

As a bridge, its vision is to link and accelerate every productivity revolution with Web3. Taking the AIGC revolution as an example, Tie Protocol’s solution to the pain points of a large number of AI gadgets on the market mentioned in this article is:

First of all, it provides developers with one-click deployment of NFT matrix, AI SDK and other basic components to help them develop and deploy AI tools and products more quickly and easily, thereby accelerating their GTM and capitalization process.

Secondly, the decentralized architecture of Tie Protocol ensures the privacy and ownership of user data, allowing users to fully control their own data and enjoy most of their value in the productivity revolution. Users can obtain, manage, and exchange their own data assets through the Tie Protocol platform, and can participate in voting, interaction, and governance through the community DAO, so that users can participate more deeply and more interestingly.

first level title