On the occasion of the release of GPT-4: How will Web3 accelerate the development of AI and the productivity revolution?
Not only the AI revolution, but every future productivity revolution or disruptive innovation can be accelerated and empowered in a native way through Web3.Original Author: Gus Liu, Researcher at Tsinghua Blockchain Association (THUBA)
Core Insights, TL;DR
- AIGC will be the most important empowering technology to enhance overall social productivity in the 21st century. The essence of Web3 is to reform production relations, returning data asset ownership to users, which is closely related to the AI revolution.
In the Web2 era, a large amount of data was controlled by internet giants, and users could only be consumers of data without ownership. In the Web3 era, blockchain technology transforms data into an encrypted, secure, and traceable asset, returning data ownership to users.
This change is crucial for the AI revolution, as data is the core driving force of AI. AI models require vast amounts of data for training and optimization. Web3 allows users to control their data ownership better, providing more diverse and comprehensive data for AI models, thereby accelerating AI development.
- A wave of AI tools is emerging, but it faces many pain points, and Web3 offers excellent solutions.
AI tools refer to small, specialized AI applications, such as text analysis tools and image recognition tools. With OpenAI opening the ChatGPT API, a large number of AI tools have begun to emerge. However, these AI tools currently face many pain points, including technical issues, market challenges, and even difficulties in capitalization. Web3 can provide solutions to these problems.
- From an investment perspective, the focus is currently on Web3 as a service protocol to empower and accelerate the productivity revolution with new native tools.
Currently, many applications or narratives in the market are still focused on: 1. Using AI to generate image NFTs with one click, continuously enriching the metaverse's scene construction; 2. X to Earn models, such as providing corrections and feedback to ChatGPT to train data and earn corresponding tokens. These applications are still in their early stages, but with continuous technological development and popularization, many more application scenarios can be explored and utilized in the future.
We believe that at this stage, investing in Web3 service protocols is superior to investing in specific toC tool products, especially service protocols aimed at new tools and new revolutions, rather than inefficiently attacking Web2 social networks.
- Not only the AI revolution, but every future productivity revolution or disruptive innovation can be accelerated and empowered through Web3's native approach (for example, Musk's helplessness in the face of GPT-4's release, hoping for Neuralink, but theoretically, Neuralink's derivative applications could also be accelerated through Web3). In fact, the best practice is to view Web3 as a component, combining productivity and changes in the real economy to unleash powerful innovative forces. In this process, Web3 no longer exists solely around financial attributes and speculation but truly plays its essence in changing production relations and empowering innovation.
1. Introduction
With the continuous development of human society, various productivity revolutions have emerged, driving social progress and development. In recent years, the rapid development of artificial intelligence technology has brought about an unprecedented revolution, a productivity revolution that encompasses almost all areas of human society. The widespread application of AI technology has not only greatly improved production efficiency and quality but has also brought about various innovative applications and business models. However, the AI revolution also faces many problems and challenges, one of the most important being how to quickly promote and implement these AI technologies, making them a part of productivity and bringing 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 more secure and autonomous network environment but also enables more open and decentralized applications, offering broader scenarios and business models, allowing AI technology to better serve various aspects of human society.
Therefore, this article will focus on "How Web3 Accelerates Every Productivity Revolution? Taking the Wave of AI Tools as an Example," introducing the basic concepts and characteristics of Web3 and AI technology, analyzing the interrelationship between the two, exploring how Web3 accelerates the development of AI tools, and looking forward to the future development prospects of Web3 and productivity revolutions.
2. Past Productivity Revolutions and Today
Past productivity revolutions can be traced back to the Industrial Revolution of the 18th century, a significant transformation in human history. Before this, the primary modes of production in human society were manual labor and agricultural production, with very low productivity. The Industrial Revolution greatly improved production efficiency and capacity through mechanized production, mass production, and factory organization. This transformation fundamentally changed the face of human society and laid the foundation for modern industrialized society.
After the Industrial Revolution, a series of productivity revolutions emerged, including the electrical revolution and the information revolution. The electrical revolution primarily mechanized and automated production further through the widespread application of electricity and the development of electrical equipment, significantly enhancing production efficiency. The information revolution mainly promoted the popularization of information production, the internet, and mobile communication technologies through the application of computer technology, making information acquisition and transmission more convenient and rapid.
These productivity revolutions not only propelled human society's progress and development but also provided tremendous driving forces for economic development and social change. Each productivity revolution represents a historical advancement, bringing new opportunities and challenges.
So, how will the AI revolution and future potential productivity revolutions differ from the past?
The core is data-driven; artificial intelligence systems require vast amounts of data for training and optimization, making data a crucial driving force for AI development. Compared to previous productivity revolutions, the AI revolution places greater emphasis on data collection, processing, and analysis. This aligns perfectly with the attributes of Web3.
3. Web3 and AI: The Combination of Two Revolutionary Forces
Web3, also known as the decentralized web, utilizes blockchain and decentralization technologies to distribute data and applications across multiple nodes, achieving a more secure, transparent, and decentralized network. The mission of Web3 is to bring about a fairer and more open internet, granting users greater control over their data and asset value.
Gus, a researcher at Tsinghua University Blockchain Association, believes that as a disruptive innovation, AI liberates productivity, while Web3 reforms production relations, returning data asset ownership to users, thereby transforming production relations. Therefore, the combination of the two holds great potential.
From a top-level design perspective, first, AI can help Web3 better achieve decentralization and autonomy. For example, AI can be used for automating contract execution and decision-making in DAOs, improving the efficiency and reliability of Web3. At the same time, Web3 technology can provide a more secure, transparent, and decentralized infrastructure for AI, making AI's learning and decision-making fairer and more trustworthy.
Secondly, the combination of Web3 and AI can bring users more intelligent and personalized applications and services. For instance, a decentralized identity management system based on Web3 can provide AI with more accurate and secure data, offering users more convenient and secure identity authentication services. Meanwhile, AI can provide personalized recommendations and services based on users' preferences and behavioral habits.
Finally, regarding specific scenarios for the combination of Web3 and AI, market voices generally focus on leveraging AI's generative capabilities to empower metaverse scenarios. For example, using a small tool that generates images with AI (integrating with Midjourney or Stable Diffusion) to create NFTs of images, text, videos, and task representations that can be placed in games and metaverse scenes for customized production. Alternatively, using AI to automatically generate questions to prevent cheating bots or witch attacks, such as Readon using ChatGPT to automatically generate quizzes to prevent users from gaming the system.
These logics essentially start from the perspective of AI tools, seeking usage scenarios in the metaverse or Web3, but undoubtedly, under this mindset, the usage scenarios of AI tools themselves are limited. Therefore, the more critical question should be from another perspective: how Web3 can empower the wave of AI tools and even extend to using Web3 to serve and accelerate every future productivity revolution. We believe that Web3 can help accelerate every productivity revolution, including AI. Through decentralization, trustlessness, and autonomy, it can promote the innovation and development of tools, solve cold start and capitalization challenges, thereby advancing the productivity revolution. This is also the key point of this article.
4. Market Perspective on the Pain Points of AI Tools
Recently (in early March 2023), after OpenAI opened the latest ChatGPT API, a large number of AI tools emerged in the market. According to Gus's incomplete statistics, within a week of OpenAI opening the latest ChatGPT API, over 1,200 AI tools appeared in the market, covering various workflows, including but not limited to: programming-related training models, code development; text processing tools, such as translation assistance, PDF summarization (ChatPDF), and automatic Excel generation (ChatExcel); as well as many fun tools, such as Xiaohongshu copywriting generators, weekly and daily report generators, etc.
With OpenAI officially launching GPT-4 in mid-March 2023, further enhancing productivity, we expect this wave of AI tool entrepreneurship to continuously emerge, exhibiting multimodal and cross-domain characteristics. Multimodal refers to AI tools being able to handle multiple data types simultaneously, such as text, images, and sound, enabling more comprehensive and precise analysis and applications. Cross-domain means that AI tools can be applied in multiple fields, such as healthcare, finance, and education, thus achieving broader application scenarios.
Undoubtedly, based on existing LLM large models, these AI tools can help us improve work efficiency, reduce costs, optimize decision-making, and even enhance our quality of life.
However, the development of these tools still faces many challenges, and we focus here on the market perspective. The business models of these tools are often quite singular, primarily relying on subscriptions or sales for profit. Survey results show that the mainstream business model for AI tools is to charge a subscription fee of $5/month (e.g., ChatPDF). The advantage of this model is its relatively low price, making it easy for users to accept, while also providing stable income for developers. However, this model also has some issues.
First, the subscription fee model has become very common in market competition, leading to intense competition where developers need to continuously optimize pricing and functionality to attract more users. This may result in some tools only offering basic functionalities, unable to provide more advanced services, making it difficult to highlight their competitive advantages.
Secondly, the market promotion of AI tools also faces challenges. Developers overly rely on personal influence on social media, lacking genuine marketing strategies, resulting in increasingly high customer acquisition costs. In such fierce market competition, merely relying on personal influence is far from sufficient to attract more users.
Finally, the narrative around AI tools is also lacking, with insufficient engaging stories and attractive brand images. This makes it difficult for users to establish a sense of identity with the tools, leading to poor brand effects in the market and difficulty in expanding market share.
In summary, AI tools face challenges in business models, market promotion, and brand building. With many similar products already in the market, competition is fierce. These tools also face significant go-to-market (GTM) difficulties, namely how to bring products to market and gain user recognition. Additionally, some tools are limited by their functionalities and narrative capabilities, unable to expand into broader markets, resulting in a low total addressable market (TAM) and insufficient capital value.
5. How Web3 Accelerates the Productivity Revolution
From a market perspective, for the AIGC revolution, using Web3 can accelerate the GTM and cold start of these small tools and products, even quickly achieving capitalization. Developers can leverage decentralized exchanges and crowdfunding platforms to help small companies and individual developers quickly raise funds, lowering financing thresholds and accelerating product launches.
Moreover, users can enjoy most of the value in the productivity revolution—data assets. Web3 adopts a decentralized model, utilizing blockchain technology to achieve decentralized storage and ownership of data assets. This allows users to own their data, rather than having it controlled by large tech companies as in the Web2 era.
At the same time, mechanisms such as DAOs and snapshot voting allow users to express their needs or vote for their favorite tools, quantifying those needs. These mechanisms enable users to participate more actively in product development and improvement, enhancing user satisfaction and engagement.
From a technical perspective, the decentralized and distributed characteristics of Web3 provide better support for AI tools. In the Web3 ecosystem, all data and applications are stored on a decentralized blockchain network, where this data is public, transparent, and immutable. This data architecture allows AI tools to more easily acquire, share, and manage data while ensuring data security and privacy. Additionally, Web3's smart contract functionality can provide AI tools with more flexible and efficient transaction, collaboration, and governance mechanisms, making the application and management of AI tools more convenient and effective.
These features can help improve the performance and reliability of AI tools. Web3 employs distributed computing, storage, and communication technologies to provide a more efficient, reliable, and secure computing and communication environment for AI tools. For example, Web3's IPFS distributed file system can offer faster, more reliable, and decentralized data storage and sharing services for AI tools, reducing data latency and loss. Web3's P2P communication protocol can also provide faster, more reliable, and decentralized communication services for AI tools, supporting real-time interaction and collaboration between AI tools.
Similarly, for Web3, this is also a rare opportunity.
From a first-principles perspective, the focus of Web3's narrative is Total Value Locked (TVL). The explosive wave of AI tools will significantly accelerate the achievement of this key metric. Currently, Web3 has relatively few application scenarios, often reusing existing Web2 narratives, such as money laundering, financial speculation, and social networking. However, these scenarios have already been occupied by Web2, making it difficult to achieve sufficient ROI by competing for these scenarios. Therefore, we need to find new opportunities and new increments. The new tools and application scenarios brought about by disruptive innovations like AIGC are precisely the new opportunities we should seize. These new tools can bring new users, traffic, and revenue sources to Web3 platforms. By integrating with AI tools, Web3 can create more useful application scenarios, providing users with better experiences and services.
6. Typical Applications of Web3 Accelerating the Productivity Revolution
As mentioned earlier, we find that many current applications or narratives in the market are still focused on:
Using AI to generate image NFTs with one click, continuously enriching the metaverse's scene construction;
X to Earn models, such as providing corrections and feedback to ChatGPT to train data and earn corresponding tokens.
These applications are still in their early stages, but with continuous technological development and popularization, many more application scenarios can be explored and utilized in the future.
We believe that at this stage, investing in Web3 service protocols is superior to investing in specific toC tool products, especially service protocols aimed at new tools and new revolutions, rather than inefficiently attacking Web2 social networks.
An article by Foresight Ventures titled "Web3 Application Layer Innovation Under the AIGC Wave" mentions an example of a Web3 service protocol aimed at achieving this goal. We will briefly introduce this product from the Tsinghua and MIT teams, Tie Protocol. As a Web3 service protocol focused on disruptive innovation technologies like AI, Tie Protocol aims to explore new Web3 native application scenarios rather than merely occupying those already claimed by Web2, thereby providing higher ROI and value. To some extent, Tie Protocol's positioning can be seen as the "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, in response to the pain points of numerous AI tools mentioned in this article, Tie Protocol's solution is:
First, it provides developers with services for one-click deployment of NFT matrices, AI SDKs, and other foundational components, helping them develop and deploy AI tools and products more quickly and easily, thus accelerating their GTM and capitalization processes.
Second, Tie Protocol's decentralized architecture ensures user data privacy and ownership, allowing users to fully control their data and enjoy most of the value in the productivity revolution. Users can acquire, manage, and exchange their data assets through the Tie Protocol platform, and participate in voting, interaction, and governance through community DAOs, making user participation deeper and more engaging.
Additionally, Tie Protocol offers smart contracts, on-chain governance, decentralized markets, and other features, providing comprehensive support and guarantees for the development, deployment, trading, and management of AI tools and products. Through these features, Tie Protocol enables developers and users to better leverage Web3 technology to accelerate the AIGC revolution, creating more commercial and social value.
7. What's Next?
Explore the infinite possibilities of the combination of Web3 and AI!
We have established a Web3 Empowering AI Talent Program, inviting all AI-related entrepreneurs, developers, and believers to join our community. Here, you can share experiences and insights with peers, discuss the applications and promotion of Web3 in AI, and accelerate the productivity revolution. In the future, we plan to develop some specific AI tools, utilizing Web3 technology for more efficient promotion and capitalization, allowing members to use Web3's voting and governance mechanisms for collective decision-making and collaboration.
8. Author Introduction
Gus Liu, Researcher at Tsinghua University Blockchain Association (THUBA), has extensive experience in the Web3 field. Twitter: @TheGusGod
