Intuition 백서 전문

bservable and are often deduced from user behavior or patterns. For example, the frequency and context of user interactions with certain data points can imply their support or belief. Implicit signal requires sophisticated analysis to interpret accurately, as it is embedded within the user’s activity data. T ransitive Signal Transitive signal refers to the trust or belief that is passed along through a network of relationships. This type of signal leverages the idea that trust can be extended through connections. For example, if User A attests to something about User B, and User C trusts User A, then User C is likely to extend some level of trust to User B, even without a direct interaction. This transitive nature of trust is vital in decentralized systems like Intuition, where direct trust relationships are not always possible. By harnessing transitive signals, the system can create a more interconnected and trust-rich environment, where the credibility of one entity can inﬂuence the perception and trustworthiness of others within the network. Signal in Intuition Though all systems naturally generate implicit signal, explicit signal in Intuition is expressed in a novel format that enables and incentivizes the creation of many-to-one, non-deterministic attestations. In Intuition, these semantic statements do not have a single ‘issuer’ - instead, anyone/anything can signal support or rejection of any existing statement/attestation at any point in time. This decoupling of information from ‘who is presently expressing/supporting that information’ introduces the ability to accommodate many-to-one attestations, making the data signiﬁcantly more usable - especially for more complex use cases that involve the concept of transitive signal, such as reputation. 43 To accomplish this, the Intuition system enables users to maintain positions on Atoms and Triples, inter- preting these positions as explicit signal from the user. This mechanism can be implemented using tokens or other accounting systems that have the ability to track user balances over time across an arbitrarily large number of entities. Atom Signal Within the Intuition framework, users signal their belief in the relevance of an Atom by adjusting their balance on that Atom. This balance can be increased or decreased arbitrarily. A balance of zero implies no signal, while any positive balance indicates a degree of belief in the Atom’s relevance. This mechanism is economically driven: users earn fees proportional to their ownership stake in an Atom as other users interact with it. Thus, if a user deems an Atom relevant, they are incentivized to signal its relevance to receive these rewards. Negative integers may be used if the Intuition system’s implementation aims to allow users to explicitly signal the irrelevance of an Atom. In systems without negative balances, the absence of a positive signal can serve as an indication of the Atom’s irrelevance. T riple Signal In the Intuition framework, users signal their belief in both the relevance and truthfulness of a Triple by modifying their balance on that Triple. Triple balances can be any integer, allowing for a nuanced expression of belief. A balance of zero indicates no signal. A negative balance signals a rejection of the Triple, explicitly indicating that the Triple is considered false, and that this falsehood is relevant. Conversely, a positive balance aﬃrms the Triple, explicitly indicating that the Triple is considered true, and that this truthfulness is relevant. For example, with a Triple asserting “[Vitalik] [is] [trustworthy],” users would express their belief in its truthfulness, and the relevancy of this truthfulness, by increasing their balance on the aﬃrmative side of the Triple, or signal disbelief by decreasing their balance into the negative. This mechanism allows for nuanced expressions of trust and belief within the decentralized ecosystem. This system enables each “statement” to exist in diﬀerent states based on who is aﬃrming or rejecting it. To illustrate the power of this approach, consider the example of building a decentralized list of followers. In a traditional one-to-one model, where each claim or attestation has a single issuer, if a user had 1,000 followers, there would be 1,000 separate “following” claims. Most of this data would be redundant, as the “following Y” part would remain constant across all claims. Additionally, this data would only be easily reconcilable assuming consistent standards and schemas across all claims. In contrast, the Intuition model simpliﬁes this by having a single Triple representing a statement such as “I am following Y. ” Users would then adjust their balances on this statement to express their stance. To see 44 all followers of Y, one would simply examine the “I am following Y” Triple and identify who has positive balances on the aﬃrmative side. This approach consolidates data, reducing redundancy and making it easier to verify and aggregate information. Furthermore, the ability to adjust balances allows for a more nuanced expression of preferences. In the context of a ‘follow’ functionality, instead of a ﬂat follow list where all followers are treated equally, users could explicitly signal varying levels of interest or endorsement. By increasing their balance on the followers they value more, users can eﬀectively rank their follow list. This means that someone could show a stronger preference for certain individuals or entities by staking more tokens on their aﬃrmations, creating a dynamic and tiered system of trust and preference. This not only enhances the granularity of data but also provides richer insights into user preferences and social dynamics within the network. 8.3.4 Fees & Economics Interactions within the Intuition system incur a fee, comparable to a gas cost in blockchain transactions. This fee serves several essential purposes: Firstly, in decentralized and permissionless systems with shared infrastructure, it is vital to prevent system abuse, such as Sybil and denial-of-service (DoS) attacks. Intuition mitigates these risks by employing an economic model similar to those in blockchain networks, necessitating a fee for data creation. This economic disincentive discourages abuse, thereby preserving the system’s integrity and functionality. Furthermore, any attacks inadvertently beneﬁt the network due to the fee payment, much like how Ethereum beneﬁts from transaction fees even when used for non-productive purposes. This mechanism ensures the ecosystem remains robust and sustainable despite potential misuse. Secondly, the creation of coherent and valuable data is often neglected, especially within the Web3 environ- ment. Providing infrastructure for generating veriﬁable data alone has proven insuﬃcient in motivating users to produce meaningful contributions. This issue is also prevalent in Web2, where the majority of users refrain from leaving reviews on platforms such as Amazon, Yelp, or Google, and rarely endorse others on LinkedIn or contribute to Wikipedia. Thus, there is a clear need for incentives to promote active and meaningful participation in the data contribution process, similar to how block rewards encourage participation in the layer 1 blockchain consensus process. Thirdly, the sheer volume of data generated globally has reached overwhelming proportions, leading to an abundance of low-quality, redundant, or irrelevant information. This overabundance dilutes the value of truly meaningful and actionable data, complicating eﬀorts to derive valuable insights. In both Web2 and Web3 environments, the emphasis needs to shift from merely producing more data to generating high-quality, reliable information. Intuition addresses this challenge by implementing mechanisms that discourage the production of irrelevant data and promote the creation of useful, pertinent information through economic incentives. By introducing an economic cost and associated rewards to data creation and curation, Intuition ensures that contributors are motivated to generate data that is coherent, valuable, and meets predeﬁned standards of relevance and accuracy. This economic model not only deters the proliferation of “junk data” but also encourages the continuous reﬁnement and validation of existing data. Fourthly, the process of establishing standards in most industries has historically been fraught with diﬃcul- ties, often described as “standards hell. ” This status quo has failed to adequately address the needs of our ecosystem. Intuition’s system of trustless economic incentives expands the concept of leveraging ﬁnancial rewards for distributed consensus—a principle successfully demonstrated in the blockchain ecosystem—to additional domains requiring social consensus and global coordination. These domains include standards for data structures, schemas, and formats, as well as canonical identiﬁers to which this data can be attached and correlated. Intuition’s imposed fees addresses these challenges in two main ways: 1. Granting Ownership in Data : A portion of the fee contributes to granting the user ownership in the data they interact with. This mechanism ensures that users have a vested interest in the data they create or engage with, promoting responsible and meaningful interactions. 2. Rewarding Data Owners : A portion of the fee is distributed to existing owners of the data being interacted with. This incentivizes the creation and maintenance of valuable data, as users receive economic rewards for their contributions to the ecosystem. 45 This ﬂow of value is enabled by Intuition’s innovative approach to data representation, which encompasses Atoms, Triples, and Signal. By structuring data into fractals via discrete, ownable fragments, this model allows for the programmatic distribution of value throughout the system’s state. Consider a user who wishes to create a new data entry stating that they like a YouTube video. The user must pay a fee to create this data, part of which grants them ownership in the statement and part of which rewards previous owners of related data. Other users who agree with this statement can also pay a fee to do so explicitly, reinforcing the validity and increasing the value of the data. This process helps to ensure that only high-quality, relevant data remains prominent, as users are ﬁnancially incentivized to support accurate and meaningful information. Figure 13: A demonstration of the high-level ﬂow, with arbitrary example numbers. This economic model encourages users to interact with data they believe will attract further engagement. Because data structures in Intuition are explicit, users are motivated to converge not only on the enti- ties/concepts/data themselves, but also on the most eﬀective ways to describe and reference these enti- ties/concepts/data. This creates an incentive for users to reach fractal consensus on data structures, from individual Atoms to complex nested Triples. By integrating these economic principles, Intuition not only secures the system against malicious attacks but also promotes a healthy, self-regulating ecosystem where users are rewarded for their contributions to the integrity and value of the data. 8.3.5 State Interpretations By implementing the aforementioned mechanisms, we achieve a state that can be interpreted in an almost limitless number of ways. Intuition deliberately avoids enforcing any speciﬁc interpretation of the data, aligning with its core principle of avoiding “truth lock-in” or the imposition of a singular perspective. The Intuition Protocol is designed to remain completely neutral regarding the interpretation of state and signal. Nevertheless, it is crucial to discuss the topic of interpretation within the Intuition system to provide an initial understanding of the various methods and perspectives that users might explore when analyzing the system’s state. Literal interpretations In the Intuition ecosystem, literal interpretations are explicit, structured expres- sions designed to facilitate coordinated, higher-order applications of its discrete state. These interpretations provide a foundation for a cohesive and interoperable framework, enabling various stakeholders to under- stand and utilize the system’s data explicitly. By achieving consensus on how speciﬁc concepts should be expressed and interpreted, the ecosystem can develop a composable system of expression and logic, similar to programming languages. 46 For instance, consider implementing an access control mechanism using the Intuition state. An implementor might use a speciﬁc Atom, deﬁned as “has access to,” as the Predicate in a Triple structure. The literal interpretation might state: if [Company’s Multisig] attests to this Triple being True, this signiﬁes that the Subject of the Triple has access to the Object—in this case, System X. This state can be used by the implementor to provision access to users of System X. By publishing this literal interpretation, the implementor allows others to observe and utilize this interpre- tation for various purposes. For example, another system might use this information to grant User A access to System Y based on their access to System X, to adjust their loan-to-value ratio in a ﬁnancial application, or to aﬀect their reputation in a particular context. The Intuition system facilitates the publication process of these literal interpretations, allowing stakeholders to register and share their interpretations as Atoms and/or Triples. By doing so, the system can be used recursively to create a registry of literal interpretations that promotes transparency and coordination. This registry enables the creation of programmatic languages on top of the Intuition protocol, leveraging the principles of discrete objects and their relationships to enforce logic and code. Algorithms In the Intuition framework, algorithms function as tools to ﬁlter and interpret data, distinct from literal interpretations which resolve to deterministic logic. Given a permissionless environment where anyone can claim anything about anything, algorithms play a crucial role, enabling signal to be eﬀectively distilled from the noise and allowing users to focus on the voices they trust. Intuition rejects the idea of ‘truth lock-in,’ acknowledging that no single algorithm can be universally opti- mal. In response, Intuition promotes a diverse and permissionless ecosystem of algorithms that users and developers can select based on their speciﬁc requirements. This empowers users and developers to create, choose, and apply the most appropriate algorithms for their particular contexts. Such ﬂexibility ensures that data ﬁltering and interpretation can be customized to meet varied needs, thereby enhancing the overall utility and eﬀectiveness of the Intuition framework. This approach contrasts sharply with the current internet landscape, where platform interactions often ne- cessitate accepting a pre-set, opaque algorithm. Intuition envisions decoupling algorithms from applications, allowing individuals to use their preferred algorithms across diﬀerent platforms - a ﬂexibility made possible through interoperable and composable data. For instance, consider an algorithm that weights and ﬁlters data based on social graphs. In this model, a user who is one degree of separation away from the observer might have their signal weighted at 100%; at two degrees away, at 80%; and at three degrees away, at 60%. This allows for nuanced data interpretation based on social proximity. Reality T unnels By decoupling each element of the data ﬁltering process, we unlock the potential to recombine these elements in novel and innovative ways. Within the Intuition framework, these combinations are known as “Reality Tunnels. ” A Reality Tunnel typically includes a mix of literal interpretations and algorithms, but it can also incorporate other relevant components essential for data ﬁltering, weighting, and sense-making. To illustrate, consider the creation of a ‘Trust Graph’ using speciﬁc Triple structures. For instance, a Triple might be formatted as ‘[Subject] [in trust graph] [Vitalik’s Web3 Trust Graph]. ’ This Trust Graph could assign diﬀerent users diﬀerent weights, explicitly signaled by a particular curator, to facilitate the weighting and ﬁltering of data. When this Trust Graph is combined with an algorithm such as EigenTrust, it can 47 Figure 14: Data ﬁltered through an algorithm, helping to distill the signal from the noise. be applied to a relevant context—like a portion of the knowledge graph pertaining to Web3 topics. This integrated viewpoint or “lens” is referred to as a Reality Tunnel. The concept of Reality Tunnels also allows users to switch between diﬀerent perspectives. Imagine engaging in a debate and having the ability to view the data from your opponent’s perspective to understand their reasoning. Currently, most platforms restrict users to a single “reality tunnel,” shaped by their social graph and a speciﬁc platform’s algorithm. However, if users could toggle between diﬀerent social graphs and apply various open-source algorithms, they would gain access to diverse viewpoints and a more comprehensive understanding of the data and the world around them. This ﬂexibility would enrich user experience by providing multiple angles on the same information, fostering better-informed discussions and decisions. 8.3.6 Intuition in Practice The introduction of Intuition’s core primitives lays a robust foundation for constructing a system that fosters a trustful interaction layer for the decentralized web. These seemingly basic primitives enable a wide range of powerful functionalities, transforming how we manage identities, data, and trust online. To illustrate this, consider, once again, the experience of buying a product on Amazon. With Intuition’s primitives, this process can be reimagined in a decentralized, trustless manner: 1. Atoms and T riples (a) Every user, product, review, and transaction can be represented as discrete objects (Atoms) and linked through relationships (Triples). For instance, a Triple could link a product (Atom) to a review (Atom) through a relationship like “hasReview. ” 2. Signal (a) Each review can be evaluated for credibility using the Signal primitive. Reviews and Signals from veriﬁed purchasers or highly reputable users can carry more weight, helping users discern trustworthy feedback from potentially fraudulent reviews. 3. F ees, Economics, and Incentives (a) Users are encouraged to express themselves, and reviewers are incentivized to provide honest and detailed feedback. Users earn money for their valuable data contributions, and their reputation becomes cross-contextual, increasing its signiﬁcance beyond a single platform. For example, a user’s status as a trusted reviewer on Amazon could serve as a valuable reputation data point in other contexts, such as securing better loan-to-value ratios in decentralized ﬁnance (DeFi) or enhancing the ability to obtain a loan. This cross-contextual reputation system contrasts with Amazon’s current reviewer reputation, which oﬀers no tangible rewards. (b) Additionally, users might be more inclined to attest to useful information about products on Amazon because their claims can be easily seen by the people they care about, thanks to portable social and trust graphs. Unlike the current Amazon platform, which lacks a friends 48 list, Intuition allows users to take their singular social graph or a set of trust graphs with them anywhere they go. This portability means that the review of one trusted individual in a speciﬁc context could hold more weight than numerous anonymous reviews. The ability to leverage these portable social and trust graphs ensures that valuable insights from trusted sources are prioritized, further enhancing the reliability and utility of the data within the ecosystem. 4. Interpretations (a) The listing of a product on a marketplac