Economic Evolution of the Web from Web1.0 to Web3

From P2P Foundation
Jump to navigation Jump to search


From Tiberius Brastaviceanu and OVN wiki [1]:

Web 1.0 and its economic reality

"Web 1.0 was built on a centralized economic model where static web pages with limited interactivity were the norm. Content was mainly created by a small number of people for a larger audience, and users had little opportunity for interactive applications. The economic model was based on the production of static web pages and did not involve the widespread user engagement and interactivity that characterize the later web versions. This model did not incorporate user-generated content, social networking, or the dynamic, interactive nature of the modern web. During the era of Web 1.0, the infrastructure was primarily funded and built by private companies, organizations, and government agencies. These entities invested in the development of the Internet's foundational technologies, such as networking equipment, servers, and communication protocols. The cost of the infrastructure for Web 1.0 was relatively trivial compared to the more expensive Web 2.0. Key players in the digital payments industry during Web 1.0 included financial institutions, credit card companies, and online payment platforms, which played a significant role in facilitating digital payments. This era was also marked by a charging model where users had to pay for each page they viewed, including directories that enabled them to find specific information. Netscape (browser) and Yahoo (search) are representative of this period.

Web 2.0 and its economic reality

The Web 2.0 iteration of the Web is characterized by increased user interaction, enabling the production of user-generated content and facilitating greater interactivity and usability for end-users. The advent of Web 2.0 has given rise to numerous social media platforms, such as Facebook, Twitter, and YouTube, that allow users to upload content and receive feedback from other users. The widespread adoption of mobile devices, such as iPhones and Android smartphones, has also contributed to the popularity of Web 2.0 through the use of apps such as WhatsApp, Instagram, Uber, and Paytm. The growth in the number of users and the increase in user-generated content has led to the transformation of the Web from a mode of communication and information gathering to a platform for software applications and e-commerce. As a result, Web 2.0 has become known as the "web as a platform". It is the Internet where we are now. Internet technology companies offered better software and services than open protocols. Users started to move away from open protocols such as the World Wide Web to centralized service platforms that could provide free services. Web 2.0 is built on the economic model of user-generated content, social networking, and dynamic, interactive web applications. This model allows individuals to collectively share resources to meet common goals, whether knowledge-based or financial, and has led to improved communication, collaboration, and knowledge sharing. The economic model of Web 2.0 also involves the use of advanced analytics and artificial intelligence. Since Web 2.0 companies can own and share user-generated content, they empower users to interact with each other by posting, liking, commenting, and other actions. Users are both creators and consumers of content in the era. The economic reality of web 2.0 has allowed the emergence at scale of a p2p economy, also called participatory economy or collaborative economy. Some technologies and platforms have been designed specifically to support p2p economic processes, such as wiki (Wikipedia) and git (Github). This p2p economy was contained within this platform economy that drove the development of web 2.0. In some sense, the p2p economy and the platform economy coexisted in a symbiotic relationship, as the development of digital commons (open source software and hardware, Wikipedia, etc.) was supported by jobs in companies and the big platforms used open source software and hardware to build, improve and maintain their infrastructure. The reality is that the p2p economy could never supersede the platform economy within the economic reality of web 2.0, because it is the platforms that contributed most to the maintenance of the Internet infrastructure.

Web 3.0, web3 and their economic reality

Web 3.0, also referred to as the Semantic Web or the Read-Write-Execute web, was seen as the next generation of the World Wide Web. It seeks to leverage artificial intelligence to make the Internet more intelligent, focusing on linking data across different platforms. This effort is guided by the World Wide Web Consortium (W3C) standards. The term "Semantic Web" was first coined by Tim Berners-Lee, who envisioned a future where machines could communicate seamlessly, creating "intelligent agents" to handle various aspects of daily life. Currently, online platforms operate as information silos, with data uploaded on one platform not automatically updating on another. The goal of the Semantic Web is to create interconnectivity between these platforms, reducing the need for redundant information uploads. Tim Berners-Lee and Ching-man Au Yeung (2009) proposed the creation of decentralized social networks that would allow users to control their data. The paper emphasizes giving users privacy and allowing them to decide whom to share information with. It also proposes storing information on a server or local computer that the user trusts, giving the user ownership of the data. It also mentions using URIs as the users’ internet identity, connecting data to applications. However, Tim Berners-Lee's proposal lacked an incentive for all parties and was not adopted by the Web 2.0 Internet giants. Also, while the idea of extending the exchange of information to the ownership of data, the idea does not cover the possibility of exchanging value across the World Wide Web. There was a lack of new technology that empowers users to exchange data or value freely without the interference or control of a trusted third party. Users will still have to rely on a platform or a third party if an exchange of value occurs. Before Bitcoin, it had never been realized because without a trusted central authority to oversee transactions, and there was no practical way to thwart three acts by malicious users: spending without authorization, spending without having enough balance, and double-spending. Bitcoin and blockchain laid the groundwork for the creation of Ethereum with smart contracts. Shortly after launching Ethereum in 2014, co-founder Dr Gavin Wood offered a different definition of Web 3.0 to Tim Berners - Lee. He advocated using blockchain to record public information, protect personal privacy and eliminate the need for trusted third parties with smart contracts. This is the first design to combine blockchain with the Internet and is the definition of Web3 accepted by the crypto industry. Industry builders and the media use the acronym "Web3" for convenience. The semantic Web 3.0, aims to improve efficiency and intelligence by connecting data across websites. In contrast, Web3 uses distributed ledger technology or blockchain, and prioritizes security and user control by giving them ownership of their data and identity. Web3+ encompasses all possible implementations of Web3, including both blockchain technology and cryptocurrency, encompasses the full range of possibilities for the future of Web3. Web 3.0 uses the solid pod to store user data and assign unique WebIDs for user identity. This Solid Pod serves as a centralized repository for the user's personal data. Each Solid user has their individual Solid Pod, which is stored on a decentralized network of servers. The data stored in a user's Solid Pod is unique to that individual and is not copied or replicated in any other Solid Pods. The data stored in a Solid Pod can be amended or updated by its owner. A hacker can potentially amend a person's data stored in a Solid Pod if they can gain unauthorized access to the user's account. It may not be possible to determine conclusively that a particular change was made by a hacker, as opposed to the user themselves or someone else with legitimate access to the data. Meanwhile, in the blockchain-based web3, users store their data immutable and securely in a cryptocurrency wallet accessible through private keys. Both approaches employ different technologies for data security. Web 3.0 uses technologies like RDF, SPARQL, OWL, and SKOS for data interchange, while Web3 relies on blockchain technology. Data in Web3 is secured due to its decentralized nature and difficulty in modifying or deleting data stored across multiple nodes. In contrast, data in the Solid Pod is stored at a single pod and can be altered more easily. Additionally, the keys stored in crypto wallets give users access to decentralized data stored on the blockchain. We can ask the questions, can AI become the core of the economic engine of the web? Is the economic foundation of the web shifting again with the advent of AI, and what role is AI going to play in it? Tibi thinks that the economy of web3+ will be AI-driven. Currently, we can see two tendencies, two models under development. The first model is in continuation with the platform economy. GAFA corporations possess the means of production of AI, the technical knowledge, the hardware and the data. They are training AI models on the data that they possess and offer AI as a service. The second model is just a jump ahead, but preserving key features of platform capitalism. Some organizations like SingularityNET are blending web3 (Internet plus blockchain) with AI to offer AI as a service, but decentralizing the infrastructure (computational power) and the sources of data. In other words, instead of relying on centralized data centres like Google, they crowdsource computing power like Bitcoin, allowing anyone to own machines that will be used to train AI models. Moreover, instead of relying on central pools of data, i.e. clouds, like Google, they rely on data supplied by users through their daily activities. Moreover, SingularityNET also proposes a marketplace of AT applications, allowing anyone to host specialized AI models, trained on specialized data, with the possibility of all these AI models to exchanges among themselves and enrich each other. This is called p2p or collaborative AI, not in the sense of human machine collaboration, but in the sense AI agent to another AI agent. The end goal is still AI as a service to the rest of the population. Platforms like SingularityNET are putting forward the fact that the code is open source, that the platform is open, providing to anyone the possibility to experiment with new AI models, host and train new AI models while benefiting from other AI models. We see this as way to crowdsource risk and effort to find practical AI applications while preserving the ability of aspirating everything that works into a more general AI model, making the little guy obsolete, similar to what Amazon is doing with outsourcing market discovery to platform members and producing everything that has shown success, out-competing the little guy in the end. In both of these cases, be it the platform or decentralized AI, the AI if offered as a service. The AI service will become the main economic engine of the Internet in this context, i.e. people will pay for AI assistance. This has important consequences on the p2pc culture, because it shifts the focus of Internet users away from each other to a service."

The next step: AI+blockchain-driven web and p2p

"Tibi sees a negative impact on the p2p economy in the shift from a platform-driven Internet to an AI driven Internet. That is because the platform-driven Internet requires users to pay attention to other users, to engage with other users, to produce for other users. This is part of the economic model of platforms, one produces a video on Youtube for other people. Moreover, when we have a problem or a question we go on the Internet to find a solution or an answer, but in doing that we are essentially going to see what others (other humans, our peers), have proposed as a solution or answer. We go on Youtube and search a video made by someone else that presents a solution, or go to Wikipedia to see what the consensus is on the topic. When we go on the Internet we essentially go to find others, engage with others, intermediated by the platform. The platform only provides what others have made, it is an intermediary between us and the rest of us. With AI as a service, when we have a problem we ask an AI agent. The web interface becomes an AI prompt, not a list of things purposed by other people. We are not engaging with our peers in our quest for answers or solutions, we are engaging with an abstract being, the AI agent. Tibi believes that this AI-driven Internet expunges social relations from our interaction with the web, turning every one of us inward, as we don't need to pay attention to others, we are not creating content for others, we are not consuming other people's content, out attention shifts to the AI, as an omniscient and omnipresent being. The best metaphor for considering AI in this context is a divinity, God, and it will suck al out attention away fro the other human beings.

Fortunately, this is not the only force exerted on the p2p economy. Hugo Mathecowitsch, founder of Tools for Commons mentions in an interview:

We must create the institution of truth, so that we can ultimately verify History. Blockchain is this institution. “Don’t trust -> verify!”. If the news is fake, imagine History!; History is a cryptic epic twisting trajectory told by the winners. The institution of truth was The Church/God; then Science; then the Press (Google / Facebook in its current state). Which now be the blockchain -> "the network / P2P". We need to be able to verify History to have a reliable database from which to work and apply AI. see video between 15-20 min.

The first thing to realize here is that AI models are only good if they are trained on good data. Good data can also mean verifiable data, or information that represents a consensus or factual information, etc. Hugo is saying that society has given itself different validation mechanisms to establish truth. For example, the press came with the journalistic profession, which imposed strict ethics for verifying the information and the source before publishing (got corrupted). Science also prescribes its validation mechanisms, which is the requirement to construct scientific theories that can be (necessarily) invalidated, and the reproducibility of experiments. Blockchain (a technology produced by the p2p culture) has at its core a consensus mechanism, a mechanism of validation, of cross correlation of data, with high probability of accuracy, which rests on a competitive dynamic among stakeholders. Moreover, blockchain is a chained chunks of data (blocks), making it very difficult to modify then past. This is known as immutability of data. Thus, everything that goes in a block is cross-correlated (verified) and once in the block it becomes almost impossible to modify. That works well for simple datasets, such as token transactions. This scheme can be extended to proof of location (knowing where something is by cross-corellating competing geolocating nodes), proof of temperature (by cross correlating competing temperature sensor nodes), etc. It is much more difficult to apply the same technology to more complex datasets or information. Thus, Hugo's proposition is very interesting but limited in capabilities. In other words, the blockchain can be used as a predictor of weather based on environmental data produced by a network of sensor nodes that also have validation capabilities, dumping data into a chained database. We can trust this AI-based prediction model because we can trust the data. But it is hard to imagine that the same AI-blockchain combination van also be used to predict for example social trends, since there are no sound blockchain-based validation mechanisms for information about people's social preferences. The blockchain technology deals well with quantities, not with qualities. But the idea that p2p as a culture, as a way of organizing social production can become a source of validated, cross-correlated information is a good one. Think about science, it relies on peer review for validation, i.e. it uses p2p mechanisms to establish scientific truth.

The p2p culture strives to eliminate hubs or the accumulation of influence in processes. It strives to distribute key processes, relaying on stigmergy, cross-correlation, peer validation, ... If blockchain can constitute the basis on AI agents as prediction models for quantitative processes, the p2p culture as a whole can become the basis of general AI, to assist humans in their actions.

Therefore, as much as an AI-based Internet is oriented towards a service model, influenced by the current market economy, and in this form it is eroding the p2p culture, it also relies on p2p to produce valid information to be fed into AI prediction models. We also need to understand the trend of the p2p economy to move away from monetary currency and the market, a post monetary, post market economy, in order to envisage a shift in the economic model of an AI-driven Internet, where AI is not build as a service but as a commons".


More information

  • Sources: Understanding the Evolution of the Internet: Web1.0 to Web3.0, Web3 and Web 3 plus