An Overview of Eight Types of Information Games Driven by Cryptocurrency

Original Title: 《Crypto-Powered Information Games》

*Author: Benjamin Funk, *Archetype

Compiled by: Elvin, ChainCatcher

Our brains, books, and databases are both recipients and creators of humanity's ever-growing tendency to generate data. The latest addition to this long list—the internet—generates and stores about 2.5 trillion bytes of data every day. While it’s easy to be awed by this number, the data points themselves hold little value. They resemble a vast, scattered puzzle that requires careful collection, processing, and contextual integration to become valuable information.

Many of today’s internet giants center their entire business models around this process, with few companies doing it as successfully as Google. Their process is as follows: they extract vast amounts of valuable raw material—"digital exhaust" generated by billions of people in the form of private data—and channel it through proprietary algorithms to predict the choices individuals might make. The more data Google extracts and processes about us, the better insights they can provide to advertisers, who in turn bid more in Google’s ad auctions, trying to convert us into customers.

The result of these processes is that Google generates $240 billion in ad revenue each year.

While Google intentionally seeks to exclude humans from this process, there is another way to generate more powerful, valuable information and monetize it—by involving humans as players in games centered around our intrinsic desire to create, search for, and speculate on information. From sports betting to MEV, to social deduction games like Among Us, we are naturally drawn to "information games" that require us to cleverly hide and discover information.

Some information games are just that. But as we will see, others can be used to generate new, valuable information and profit from it, serving as the backbone of a new generation of products and business models.

However, information games always have a fatal flaw: trust. Specifically, players need to trust that other players will not share information or act on it in ways that violate the rules of the game. If players in Among Us could switch from teammates to impostors midway through the game, or if block builders could compute incorrect state roots and still be accepted by validators, no one would want to play anymore. To address this trust issue, we turn to trusted third parties to create and mediate information games for us.

This works well for low-stakes games like Among Us, but limiting the creation and mediation of games to a centralized party constrains the trust and experimentation around the types of information games we can play, thereby limiting the types of information we can collect, utilize, and monetize.

In short, there are many information games that have yet to be attempted because we have not found a way to keep them fair and trustworthy in a decentralized environment.

Programmable blockchains and new cryptographic primitives are addressing this issue, allowing us to create and coordinate information games at scale without permission, without having to trust third parties or each other.

In turn, crypto-powered information games can rapidly increase the quantity and quality of information available in the world, enhance our collective decision-making capabilities, and unleash efficiency gains on a global GDP scale. Imagine globally accessible prediction markets as tools for allocating capital to internet-native large funds. Or a game that allows individuals to pool their private health data and be rewarded for any new discoveries arising from the use of that data, all while protecting their privacy.

However, as this article will show, crypto-centric information games may not yet be ready to tackle these high-stakes use cases. But by experimenting with smaller, engaging information games today, teams can focus on attracting players and building trust before potentially scaling up to create and monetize more profitable information markets tomorrow.

From prediction markets to game-theoretic oracles and TEE networks, this article will cover the design space for creating these crypto-powered information games and the key infrastructure needed to unlock their potential.

Permissionless Markets: Prerequisites for Information Games

From the future to information markets, blockchains allow developers to create customizable automated financial devices that underpin permissionless, unstoppable markets. Thus, anyone can now create mechanisms to incentivize, coordinate, and resolve value and information exchanges. This underscores the critical role of blockchains in enabling us to rapidly experiment with how to best configure games to deliver maximum value to each participant.

It is very difficult to persuade centralized intermediaries to adapt at this speed or allow their users to participate in these experiments. Therefore, permissionless markets will become the medium through which fringe theories and cutting-edge research papers can be realized. We have already seen this happen in the context of prediction markets, where theoretical automated market-making strategies conceptualized to address the low liquidity of prediction markets have been implemented on crypto rails in the form of CPMM and tested with real money.

Permissionless markets are significant drivers of better information generation and monetization.

Information-Producing Information Games

Many information games generate new information for players to make better decisions.

These information games create incentive mechanisms to extract raw materials (public and private data) from people, databases, and other sources, and then aggregate this data through the best information production machines (markets and algorithms). Ideally, new information can be generated and monetized by helping other players make the right decisions while aggregating this information. For example, investment DAOs use the results of prediction markets to determine whether to invest in new startups.

The games and tools used by information game designers vary based on the types of information they might produce, and we have a vast design space with different challenges and opportunities to explore.

But let’s start with the most actively developed and discussed information game today—prediction markets.

Game #1: Prediction Markets as Tools for Generating Information

One of the most popular information games we see in cryptocurrency (and beyond) is prediction markets. Polymarket is the world’s leading prediction market, leveraging the crypto rails to achieve over $400 million in cumulative trading volume (and growing rapidly).

Prediction markets work by incentivizing players to bet their own capital (or game tokens) on the outcomes of various events. This requirement for personal economic interest or "skin in the game" helps ensure that participants are genuinely committed to their predictions. As traders act on their insights by buying undervalued shares and selling overvalued ones, the market dynamically adjusts. These adjustments in market prices reflect a more accurate collective estimate of event probabilities, effectively correcting any initial mispricings.

The more people with different but relevant public and private knowledge bet on the market, the more the prices can reflect the truth. Ultimately, prediction markets drive the accurate aggregation of information by leveraging financial incentives, thus harnessing the "wisdom of the crowd."

Unfortunately, prediction markets face several key challenges, many of which boil down to various scalability issues.

Truth Bottleneck

The Keynesian beauty contest—a game where judges aim to select the option they believe other judges will also select—is not unique to prediction markets. However, its negative impact is more pronounced here because the goal of prediction markets is to create accurate information. Moreover, unlike traditional financial markets, where profit maximization primarily drives participant behavior, bettors in prediction markets are more likely to be influenced by personal beliefs, political biases, or vested interests in certain outcomes. Therefore, if their bets resonate with their personal values or expectations of profit from behavior outside the market, they may be more willing to suffer financial losses in the market itself.

Additionally, the more people view any market or algorithm as a source of truth, the higher the motivation to manipulate that market. This is not much different from the problems faced by social media. The more people trust the information products generated by social media platforms, the higher the motivation to manipulate them for profit or socio-political gain.

Some participants may even exploit the signals and incentives generated by prediction markets to reprice collective beliefs and encourage collective action. For example, imagine a government using some form of "quantitative easing" to influence prediction markets on key issues like climate change or war. By purchasing large amounts of shares in relevant prediction markets, they could shift financial incentives toward desired outcomes. Perhaps they have determined that the systemic risks of climate change are underestimated, so they buy a large amount of "no" shares in the market predicting climate improvement by 2028. This action could encourage more climate startups to develop technologies, giving them an informational edge to bet on "yes" shares, thereby accelerating efforts to find solutions.

While the above factors have been shown to negatively impact the quality of information produced, there is also evidence that manipulative behavior can actually enhance market accuracy, as market manipulators are noise traders who inform market participants that they can profit through trading.

Thus, we can infer that the aforementioned issues stem from a lack of sufficient capitalized, informed traders to help correct the market. Allowing these informed traders to borrow and short-sell may be a key means to improve the efficiency of these markets.

Moreover, in longer-term markets, it becomes more difficult for informed traders to resist manipulation, as manipulators have more time to reflexively influence market sentiment and actual outcomes through trading. Implementing shorter, more frequent resolution dates for markets can enhance trust in the game (thereby improving the quality of its information) while also making gameplay more engaging.

We have also seen early signs that, in some cases, players prefer manipulable information games with market resolutions. Perl is the top account on Farcaster, and at the time of writing, it has adopted this model and created an in-app platform to speculate on user engagement. Prediction markets like "Will @ace or @dwr.eth (co-founders of Perl and Farcaster, respectively) get more likes tomorrow?" once initiated, lead to teams and their fans starting to meme. Only here, the game is played asynchronously and measured by likes instead of touchdowns. While Perl's game intentionally subverts the quality of information generation in prediction markets, an interesting meta-game emerges by coordinating resolutions that favor one side.

Prediction-based games can reduce manipulation and boredom by using shorter, possibly more frequent rounds. However, in low-stakes games, allowing players to manipulate can increase enjoyment and become an integral part of the gameplay.

Finding the Right Judges and Oracles

Another challenge for prediction markets lies in adjudication—how to correctly adjudicate the market? In many cases, we can rely on oracles protected by reputation and collateral, which can insert off-chain data sources. To address this issue, prediction market designers can rely on game theory and cryptographic oracles to insert broader themes, including players' private information.

Game-theoretic oracles or Schelling-point oracles assume that, in the absence of direct communication, participants (or nodes) in the network will independently converge on a single answer or outcome they believe others will also choose. These oracles were pioneered by Augur and later by UMA, encouraging honest reporting and preventing collusion by rewarding participants based on their proximity to the "consensus" answer.

Nevertheless, making these oracles reliably adjudicate bets from a small number of players still presents many challenges, with identifying and communicating with each other to collude being a potential threat. While cryptography is touted as a key tool to avoid collusion among voters, it can also be used to facilitate collusion and prevent prediction markets from resolving correctly. We can see this potential for programmatic bribery and coordinated price manipulation through DarkDAO using trusted execution environments (TEE). Blocksense is one of the teams dedicated to balancing these incentives, using secret committees and encrypted voting to prevent collusion and bribery.

Source: Hacking Distributed

Oracles challenges can also be addressed by leveraging on-chain data. In MetaDAO, players are rewarded if they correctly predict how a specific proposal will affect the price of its native token. This price is provided by UniswapV3 positions, serving as an oracle for token value.

Even so, these oracles are still limited in resolving market issues based on public data. If we could resolve market issues based on private data, we could unlock entirely new types of prediction markets.

One way we can solve market issues based on private information is to use the results of the information games themselves as oracles. Bayesian markets are one such example, relying on the principles of Bayesian inference to derive participants' beliefs about their private information by allowing bettors to bet on others' beliefs. For instance, setting up a market where people bet "How many people are satisfied with their lives?" reveals bettors' beliefs about the satisfaction levels of others. Thus, we can draw accurate conclusions about players' private information that would otherwise be unverifiable facts.

Another solution we can draw from is to utilize oracles that cleverly import data from private web2 APIs. Some existing oracles are displayed in the "Public and Private Information Oracles" section of the market map. Using these oracles, prediction markets can be created around some players' private information, incentivizing holders of private information to resolve specific prediction markets in a verifiable manner in exchange for transaction fees from bettors. More generally, the ability to securely access richer on-chain and off-chain data about people can serve as an identity primitive, helping us better identify, incentivize, and more effectively match players in information games, guiding us to the necessary information to make information games relevant to players.

Innovations in oracle design will expand the range of data we can use to resolve prediction markets, broadening the design space for information games around private information.

Liquidity Bottleneck

Attracting liquidity to prediction markets is challenging. First, these markets are binary markets where players bet "yes" or "no" on specific topics, either winning a fixed amount or winning nothing at all. Thus, the value of these shares can fluctuate dramatically with minor changes in the underlying asset price, especially as expiration approaches. This makes predicting their short-term price movements crucial yet challenging. To mitigate the significant risks of these sudden changes, traders must employ sophisticated and constantly adjusting strategies to guard against unexpected market volatility.

More importantly, as prediction markets expand their scope to more topics and extend their timeframes, attracting liquidity becomes even more difficult. The more types of markets beyond politics and sports, and the longer their duration, the less people feel they have an edge in betting on these markets. As a result, the number of bettors decreases, leading to a decline in the quality of information produced.

Prediction markets inherently face these liquidity issues because forming prices requires revealing private information and betting based on that information, both of which are costly activities. Participants need to be compensated for the effort and risks they undertake, including the costs of gathering information and locking up capital. This compensation typically comes from others willing to accept worse odds, for reasons including entertainment (i.e., sports betting) or hedging risks (i.e., oil futures), which helps drive significant liquidity and trading volume. However, prediction market topics with narrower interests have less commercial appeal to players, resulting in less liquidity and trading volume.

Economic Improvements: Stacking and Diversification

We can strive to address these issues by recycling ideas from traditional finance and other existing information games.

Notably, we can leverage the overlay concept covered by Hasu in "The Problem of Prediction Markets." In gambling tournaments, the concept of overlay (additional value added to the prize pool by the house to encourage participation) serves a similar purpose to subsidies proposed for prediction markets. Overlay effectively lowers the entry cost for players, making tournaments more attractive and increasing participation from both novices and experienced players.

Just as overlay in gambling tournaments promotes player participation by enhancing potential returns on investment, subsidies in prediction markets can incentivize participants by lowering entry barriers and increasing the economic appeal of participation. Subsidies also act as lighthouses, attracting diverse viewpoints and insights from uninformed and informed traders, who can profit by correcting these viewpoints and insights. Teams implementing this strategy must systematically identify and reach out to potential subsidy providers and create markets around their needs, as they are the ones willing to provide the necessary liquidity.

Similarly, a fund-like structure can be implemented to achieve temporal and sectoral diversification, increasing the liquidity of prediction markets across broader issues and timeframes. For example, many companies may find value in markets centered around how specific lawsuits will resolve. These companies can lower their participation costs by providing funding to legal experts, allowing them to diversify across a wide market and then rewarding them based on their long-term performance.

In this setup, traders will be able to borrow to market-make, with amounts parameterized based on the demand for the information that will be generated and the trader's reputation on that topic. This can be combined with management fees as an additional overlay for each market.

For liquidity providers, they will gain exposure to traders incentivized to bet correctly in these markets while diversifying their investments across a basket of unrelated assets with different timeframes. While agency problems must be considered, this system can increase both the scale of liquidity provided in these markets and the diversity of the funding pools allocated for liquidity. As a reward, the quality and variety of information products can improve, while also creating new information about the skills and knowledge of different market traders, accelerating returns for liquidity providers through reputational byproducts.

When the value of the information players can generate is significant, integrating composable financial markets like borrowing and liquidity mining into gameplay can serve as a key tool for lowering entry barriers.

User Experience Improvements: Simpler Interfaces and Flexible Incentives

The default, exchange-centered user experience and limited types of rewards in today’s prediction markets may drive away those motivated by other types of interfaces and incentives, further constraining liquidity. In terms of betting players, there are many interesting ways to enhance the quality of prediction markets, all centered around expanding the coverage and accessibility of different types of players.

First, we can improve the user experience of prediction markets by integrating them into larger social platforms. Perl and Swaye demonstrate how by inserting data from Farcaster, users can relieve the cognitive burden of opening a separate app, allowing information game designers to identify and guide players into their unique markets (i.e., top participants in the game).

Source: Perl on Farcaster

There are also opportunities to experiment with expanding the range of rewards allocated to bettors and establishing more lenient requirements for the capital they invest. This might look like rewarding individuals through proofs or expanding the range of economic rewards to "in-app utilities" or rights represented by points or tokens.

While monetary incentives are crucial for the operation of prediction markets, some literature suggests that virtual currencies can create prediction markets of equal quality. In fact, this tells us that we can flexibly hypothesize which types of "risk-sharing" bettors will face risks and be forced to profit.

Additionally, different types of market mechanisms can be employed to make the user experience more survey-like, further reducing friction and lowering entry barriers. A study from Cambridge University evaluated this hypothesis and found that during periods of low trading activity, wide bid-ask spreads, and rapid market resolutions, survey mechanisms can yield more accurate results compared to prediction markets. The study also found that combining survey-based prediction games with monetary incentives from prediction markets produced significantly higher accuracy than prediction market prices alone. Furthermore, to address potential challenges of information stagnation, surveys can be regularly "updated" based on some push or pull system, incentivizing dynamic re-creation based on new information.

Crypto information games have often prevented all but the most focused advanced users from participating. Now, with reduced costs, increased availability, and richer data, we have the opportunity to develop more diverse and accessible games targeted at specific audiences.

Game #2: Generating Information Through Privacy-Preserving Computation

Imagine a game for Solidity developers where players use multi-party computation (MPC) to reveal their salaries and compute an average while keeping individual salaries confidential. For cryptocurrency professionals, this would be a valuable way to negotiate with their respective employers while also serving as a source of entertainment.

More broadly, information games can leverage privacy-preserving technologies to expand the range of raw materials, particularly private data and information that can be analyzed to generate new insights. By ensuring privacy, these tools can increase the diversity and inclination of people to share data and information, compensating these data providers for the value generated.

While this is not exhaustive, some of the tools information producers use to execute this include zero-knowledge (ZK), multi-party computation (MPC), fully homomorphic encryption (FHE), and trusted execution environments (TEEs). These technologies have different core mechanisms, but they all achieve a similar goal—enabling individuals to provide sensitive information in a privacy-protecting manner.

Nevertheless, using software and hardware-based cryptographic primitives in use cases requiring strong confidentiality guarantees still presents many daunting challenges, which we will discuss later.

Privacy-preserving cryptography greatly expands the design space for new information games that were previously impossible.

Game #3: Competition Between Models to Enhance Information Production

Imagine a game where data scientists compete by developing and betting on trading models for decentralized hedge funds. Then, the blockchain reaches consensus on the scores of specific models and rewards or penalizes participants based on the accuracy of model predictions and their impact on fund returns. This is the approach taken by one of the earliest information games on Ethereum, Numerai. In this game, Ethereum's consensus is leveraged by the global competition between different models and their creators, effectively incentivizing artificial intelligence to play information games, producing valuable returns.

Going further, we can also incentivize artificial intelligence to play information games for us more directly, leveraging their encyclopedic knowledge to compete in making predictions. While they may not necessarily play these games, using intelligent machines instead of humans would significantly reduce the labor costs required to produce information. Thus, these AI models could increase the liquidity of more niche prediction markets that humans would otherwise be unwilling to participate in. As Vitalik stated:

"If you build a market and offer a $50 liquidity subsidy, humans won’t care about bidding, but thousands of AIs will easily swarm in and make their best guesses. The incentive to do well on any one question may be small, but the incentive to create an AI that can make good predictions could be millions."

Alternatively, we could create competition among machine learning models based on the value of the information they create. Teams like Allora and Bittensor TAO are working to coordinate models and agents to broadcast their predictions to others in the network, while others are responsible for evaluating, scoring, and broadcasting their performance back to the network. In each period, the collective assessment among models is used to allocate rewards or rights based on the quality of different models' predictions. Thus, entrepreneurs can leverage a self-improving network of models to enhance the quality of information flowing through their markets.

It is entirely possible for such information markets to exist: the use of models leads to information products of quality that human-to-human information games cannot compare to.

Monetizing Information Games

Some information games can sustain themselves purely through the enjoyment users derive from them. But for those looking to monetize the value of the information they generate, things require a bit more thought. Unfortunately, the quality of information as a commodity leads to severe market failures, hindering its seamless monetization:

• Information only has value after it has been consumed, making it difficult for buyers to assess whether the seller's price accurately reflects the value of their information.
• Information is non-competitive—its consumption does not diminish its availability, meaning it lacks the scarcity attributes that would interest buyers.
• Despite high initial production costs, the non-exclusivity of information, combined with low replication costs, makes it difficult for sellers to prevent unauthorized access.

These economic characteristics pose challenges for both buyers and sellers to profit from information, potentially leading to underproduction of information. If information is quickly understood by everyone who can utilize it simultaneously, the opportunities for information buyers to exploit information asymmetries will diminish due to increased competition or the collapse of the schemes they intend to use. Fortunately, there are some crypto tools available to address these issues, and they already exist.

Game #4: Exchanges—Monetizing Through Information Speculation

One way to monetize information production without confidentiality or restrictions on the range of actions that can be taken is simply to make the information public while creating a tool for people to bet on how it will change—also known as derivatives.

Parcl is a company actively engaged in this business, allowing users to speculate on the ups and downs of the real estate market. Parcl's market is supported by real-time price information, which Parcl Labs sources from a vast real estate database and provides through proprietary algorithms to generate granular, accurate information that surpasses the quality of traditional real estate price indices.

While Parcl does monetize this information more directly through APIs, they have created an additional layer of monetization by allowing traders to bet on how this information will change over time. Other projects, such as IKB and Fantasy, mentioned in the "Alternative Information Markets" section of the market map, focus on monetizing by guessing or hedging how existing public information will change (from player performance to creators' social engagement).

If you can sell the right to speculate on the information you generate, you can monetize it without confidentiality or restrictions on how buyers can use that information.

Game #5: Markets for Discovering Confidential Information

Imagine a game that allows you to discover selected alpha before the latest on-chain activities and new crypto startups are known to the world. For this, the information needs to be confidential to address the non-competitive and exclusivity issues posed by public information. Thus, the next generation of information markets is facilitating the exchange of confidential information while leveraging blockchain to discover and regulate access for all participants who can pay to access that information.

Freatic's decentralized confidential information market, Murmur, exemplifies this approach by restricting exclusive access to information through NFTs and a queue system. Information buyers first subscribe to specific topics by purchasing NFTs presented in coupon form. This way, they secure a position in the queue to redeem confidential information from publishers, paying an additional fee that allows them to slow its dissemination. Buyers can also vote on the quality of the information afterward. Through this process, Murmur ensures the confidentiality and value of information without restricting its sale to a single entity.

In contrast, Friend.tech manages access to confidential information in group chats using keys and bonding curves, with access costs increasing as demand rises. Thus, people can view Friend.tech keys as proxies for the average value of personal information (assuming the key market is effective). However, players always "price" some concept of that person's "value" when trading keys, making it difficult for buyers to price the value of the information. Perhaps this can serve as another data point to support the notion that the most valuable "information market" to date is actually the meme coin market, which, if you look closely, can act as a prediction market for the symbolic value of specific trends or figures.

Setting aside meme coins, one direction teams controlling information access can pursue is allowing information sellers to design bonding curves that better link access prices to information value. For example, for information that rapidly loses value over time, its pricing could be determined by reflecting the rapid depreciation of information value over time through a bonding curve.

Decentralized currency exchanges face challenges due to trust issues and the dual coincidence of demand. Blockchains have solved the money problem (Bitcoin) and will similarly address the information aspect, catalyzed by interesting games centered around discovering hidden information.

Game #6: Futarchy—Monetizing Prediction Markets

One primary way to monetize information without explicit confidentiality is to produce and sell information that only one organization can and will use. This strategy is not new, as many companies have already monetized information by restricting access to specific buyers through auctions or confidentiality agreements. However, we are seeing a new business model for selling information products—generating public information that is relevant and valuable only to organizations making specific decisions.

In fact, we are only now seeing prediction markets built on crypto rails to attempt to use Futarchy as an alternative mechanism for monetizing the information they generate.

Futarchy offers a novel approach to improving decision-making, centered around the information created by prediction markets. The information produced by prediction markets is used to make decisions, and when the prediction markets resolve, the best predictors are rewarded.

In and of itself, prediction markets are zero-sum games for participants, limiting the incentives for informed traders to engage and exacerbating existing liquidity bottlenecks. Futarchy can address this issue because the wealth created by better decisions can be redistributed to traders.

Crypto-native entities like MetaDAO are already experimenting with Futarchy. When a proposal is put forward, such as Pantera's proposal to purchase MetaDAO governance tokens, two prediction markets are created: "pass" indicating support and "fail" indicating opposition. Participants trade conditional tokens within these markets, speculating on the proposal's impact on DAO value. The resolution depends on the time-weighted average price (TWAP) comparison of "pass" and "fail" tokens after a specified period. If the TWAP of the "pass" market exceeds the TWAP of the "fail" market by a certain margin, the proposal is approved, executing its terms and canceling trades in the fail market. This system leverages market dynamics to drive governance decisions, aligning them with the collective predictions of the proposal's impact on increasing or decreasing DAO value.

In some cases, Futarchy must be designed around confidentiality. For example, if prediction markets are used to determine hiring decisions for specific individuals, that information would become public and turn into an information hazard—competitors might be interested in poaching employees based on market predictions.

Another reason for keeping information confidential is its impact on motivation and organizational culture. As Robin Hanson pointed out in his talk on the future of prediction markets, Google's own internal experiments faced resistance because executives were concerned that public performance metrics might demotivate employees. Of course, managers are reluctant to implement things that might expose "the emperor has no clothes," and we see this in practice today. According to MetaDAO founder @metaproph3t, some individuals have chosen not to submit proposals because they do not want to face market evaluations.

Both issues can be addressed by restricting specific decision-makers' access to prediction market information. However, by granting these decision-makers autonomy to act based on that information, bettors will incorporate these biases into their bets, thereby reducing the quality of the information generated.

In other cases, Futarchy may be better suited for specific industries where its advantages outweigh cultural impacts, such as hedge funds like Bridgewater. Integrating blockchain can further enhance the integrity of Futarchy to prevent manipulation (looking at you, Ray Dalio).

So far, prediction markets have been limited to monetization through allowing speculation or hedging. In helping organizations make better decisions, prediction markets can unlock an entirely new market, although the role of confidentiality around information remains an unresolved issue.

Game #7: Trusted Commitments in Programmable Information Games

As mentioned at the beginning of this article, Google monetizes information by renting its use to advertisers while restricting that use to Google's ad auctions. Similarly, trusted commitments help information sellers profit by limiting the actions buyers can take based on the information provided.

Information sellers can use cryptographic methods like MPC, TEE, and FHE to ensure that buyers make trusted commitments regarding the computations performed on private data. Thus, sellers can delegate their information to buyers, allowing them specific control over future actions taken around their private information without revealing the information itself.

This primitive can unlock various information games. Imagine that traders (information sellers) can only sell the right to order trades to information buyers (seekers) if the buyers commit to simulating their trading sequences up to a capped number of times. Further, imagine allowing Netflix users to delegate the right to watch Netflix movies from their accounts to others, enabling them to "harvest" rewards from their accounts without disclosing their login details. In turn, buyers can extract value from the seller's private information without the seller having to deal with the challenges of selling the information itself (information is a non-competitive, non-exclusive experiential commodity).

Unlocking Google-scale monetization for today's information game designers

TEEs provide a practical option for implementing such controls today, although their confidentiality guarantees are limited. While TEEs are not suitable for protecting large assets or sensitive data, they are applicable for use cases requiring more time-limited access to confidential information, such as front-running protection. SUAVE is a project created by the Flashbots team that is building a TEE network that developers can use now, with a long-term vision of enabling application developers to find new ways to better monetize the value of their and their customers' information.

In the design of SUAVE, integrating blockchain with TEEs addresses three key TEE limitations necessary for advancing information games. First, the blockchain eliminates the trust requirement for communication between hosts and players, as either party may engage in censorship or malicious behavior. Second, the blockchain provides a secure state maintenance mechanism that prevents rollback attacks, which TEEs are susceptible to. Finally, the blockchain is crucial for ensuring the permissionless, censorship-resistant creation of TEE-based information games (SUAPP), with smart contracts, inputs, and outputs trusted by all players.

While many early information games using SUAVE are clearly MEV-centric, they have the opportunity to be applied to information games far beyond trading.

Game #8: Reputation and Zero-Knowledge Facilitate In-Game Markets

A key challenge in monetizing information is the inherent nature of information as an "experiential commodity." The value of experiential commodities is only recognized when used, complicating sellers' ability to price them in advance. When creating mechanisms to solve this problem, we can also create engaging gameplay for users. Some games primarily focus on allowing players to build reputations that distinguish them from other players, such as World of Warcraft, which can be a source of fun but also a key way for players to decide whom to collaborate with. Other games may wish to allow sellers to commit to certain intelligence (i.e., enemy locations, secret plans) without requiring them to disclose that information in advance.

To overcome this issue, information game designers can leverage cryptographic solutions like zero-knowledge proofs (ZKP) to verify the characteristics of computational information products (e.g., the validity of trading algorithms) without disclosing the actual data or code. This can be achieved by creating cryptographic commitments, timestamping them on the blockchain, and providing ZKPs of algorithm performance. However, this approach is only effective for information products whose value derives from their computational properties and can be tested on verifiable inputs.

For other types of information products, reputation and identity become crucial. Consensus mechanisms among information buyers can be utilized to establish reputations around the value of the information sellers are trying to sell.

Systems like Murmur utilize subscriber voting within exclusive windows to establish the reputation of publishers, elevating them from unverified to verified status based on community feedback. This process creates a transparent and immutable record of interactions, establishing a trustworthy reputation for sellers through tight feedback loops.

Additionally, the Erasure Bay protocol requires sellers to stake funds along with their reputation as a signal of their information's reliability. The protocol establishes a "griefing factor," allowing buyers to destroy a portion of the seller's shares if the information proves to be of low quality, thereby ensuring sellers have the incentive to provide high-quality information.

To avoid market failures and maximize sales, game designers need to provide sellers with cryptographic tools to prove the value of their information or reliable, rapid mechanisms to establish reputations around their previously sold products.

Conclusion

Information games are not new. However, before the advent of programmable blockchains, game designers could only seek permission from centralized intermediaries, and players could only engage in games mediated by trusted third parties.

Now, the significant reduction in the cost of blockchain space means that anyone can create a DAO or confidential information protocol inspired by Futarchy and insert countless tools for validation, adjudication, monetization, and more. The games we will see unlock through low participation barriers and open innovation on permissionless financial rails are unimaginable.

This article showcases the early signs and challenges of implementing a new wave of information games, as well as the potential for using cryptographic tools to address these issues. With these tools, some game designers will improve the information games we have already played, such as trading and MEV, while others will create games that were previously impossible.

Nevertheless, each of these crypto information games represents mini-games that need to be combined to form a complete game. The joy and excitement players derive from building reputations, collaborating with teams, and vying for influence within organizations are all components of a larger whole.