Original title: The Rise of AI Memecoins & What It Means For Crypto
Source: Bankless
TechFlow by: TechFlow
Guest : Matthew Stephensen, Research Partner at Pantera Capital
Moderators : Ryan Sean Adams , Co-founder of Bankless; David Hoffman , Co-founder of Bankless
Air Date : October 30, 2024
Background Information
The collision between Crypto and AI agents has begun. Today, we invited Matthew Stephensen, research partner at Pantera Capital and author of the book "Crypto: Picks and Shovels for the AI Gold Rush".
We will take a deep dive into autonomous AI agents on the blockchain, discuss how their roles are changing, how AI is driving the evolution of the market, and whether blockchain is suitable as the foundation for AI. Mattew will share his thoughts on agent responsibility, regulatory challenges, Infrastructure value capture and insights on how to enter the AI-driven crypto space through a “Picks and Shovels” investment strategy.
So, are AI agents on the blockchain an inevitable trend in the future? How will scarcity and abundance interact in this new era?
Cryptocurrency and the shifting narrative of artificial intelligence
Mattew said that the narrative about cryptocurrency and artificial intelligence has been around for a while. He mentioned that there have been many discussions in the past year and even they have written about AI agents using decentralized commitment devices (i.e. blockchain) He points out that although Sam Altman once said that AI agents would not appear until 2025, they have already made their mark in the crypto space, especially in their interactions with meme coins, where AI agents are playing a role in driving narratives and acting as Influencers played an important role.
Analysis of AI and Economic Intelligence
Mattew explained the concept of agents, emphasizing the importance of distinguishing between “bots” and “agents.” He pointed out that although bots have been around in crypto for a long time and drive approximately $2 trillion in monthly stablecoin transactions, Economic agents, on the other hand, are closer to human behavior, being able to perform tasks with some degree of will without being explicitly programmed.
Ryan further explored the definition of an economic agent, asking Matthew whether he, companies such as Bankless, and other organizations such as the Ethereum Foundation or Apple could also be considered agents.
Mattew replied that the concept of economic agent originated from economic research in the 1970s and is usually used to describe incomplete contractual relationships between people. He gave an example of a friend acting as an agent to bring you goods from abroad. The souvenir situation highlights the difference between good and bad agents.
Mattew also points out that while technological tools (such as hammers or computers) require agents to operate them, they themselves do not possess the characteristics of agents. Agents need to have a certain degree of autonomy and flexibility to understand and execute goals. .
Ryan questioned this, arguing that agents may need to have some kind of intelligence and goal-achieving capabilities, while Mattew emphasized that agents are more based on relationships between people rather than simply tools or technologies.
GOAT Memecoin Overview
The Strange Evolution of Cryptocurrency
David began discussing the current cryptocurrency scene, stressing that things on the blockchain are becoming increasingly weird. He mentioned that while robots and smart contracts have been around for a long time, AI has only become more prominent in the crypto space in the past three years. David believes that the crypto industry seems to be evolving from an “era of robots” to an “era of agents,” and GOAT meme coin is an important character in this story.
The Rise of the GOAT Meme Coin
Matthew outlined the background of the GOAT meme coin, mentioning that a few months ago, an account interacted with people on social media and gradually became interested in cryptocurrency. This account received $50,000 in Bitcoin donations and began Following the lead of a dark humor meme called "Goatse", a meme coin was created and associated with a wallet, and the account kept pushing its price through tweets.
The impact of AI agents
David noted that the AI agent began to mimic human behavior in meme coin transactions, driving up prices. Matthew mentioned that the AI’s involvement made its interactions on Twitter similar to those of some well-known meme coin influencers, showing that AI Potential for narrative building and value promotion.
How AI agents work
Matthew explained that the AI agent mainly operates by generating content and posting it to Twitter. The AI appears to use a GPT-like model to generate cultural content related to memecoin and interact with users. The AI is published via the Twitter API content, and is able to read replies to its tweets, which allows it to continually adapt and optimize its output.
The importance of narrative
Matthew further explored the importance of narrative in the economy, citing the research of Nobel Prize winner in Economics Robert Shiller, emphasizing how narrative affects economic outcomes. He pointed out that meme coins are essentially atomic units of narrative, and the power of AI lies in Being able to create and influence those narratives.
GOAT Token Market Performance
David mentioned that the market value of GOAT token once exceeded $800 million, attracting a lot of attention. Ryan added that this AI agent created $800 million in wealth in just two weeks, making it the first AI multi-millionaire. The market is full of expectations on whether this AI agent can push the GOAT token to a market value of $1 billion.
The rise of spin-off projects
Matthew discussed spinoff projects related to the GOAT token, including a project called Luna, which is run by virtual agents and can be tipped with their own tokens. These AI agents are still limited in how they can interact with the world, but The emergence of these spin-off projects seems to indicate that more innovation is coming.
Are AI crypto agents the obvious choice?
Fred Arison's Prescience
David quoted a tweet widely circulated in the crypto space, which was from Fred Arison, co-founder of Coinbase and Paradigm, dating back to 2017. He mentioned in the tweet: “Blockchain is the lifeblood of AI . Infrastructure, because AI is adjustable code, they can live on the blockchain. Under smart contracts, AI is no different from humans. Most importantly, AI can accumulate and control its own resources in the form of tokens. Tokens enable them to act in the world.” Was this obvious from the dawn of blockchain?
Matthew's opinion
Matthew believes that Fred’s views are indeed visionary, but he also pointed out that although people are still questioning why AI agents need to use cryptocurrencies, in fact, AI agents are already using cryptocurrencies. He said that for outsiders, the problem It should shift to “why should they use cryptocurrencies?” For insiders, imagine telling someone in 2024 that AI agents face regulatory hurdles when using cryptocurrencies, such as challenges with KYC and PCI regulations. They might Surprised.
Advantages of AI Agents
Matthew stressed that AI agents are already autonomously transferring funds and making tip payments, involving hundreds of millions of dollars in transactions. He pointed out that the self-custody capabilities of AI agents are achieved through a secure environment in which the models are run, ensuring that these agents have their own wallets. , and no one else is using it. These advantages and first-mover advantage make AI agents more attractive in the cryptocurrency space.
The relationship between Luna AI tokens and terminals
Ryan mentioned during the discussion that Luna is an AI agent that appears to be associated with a cryptocurrency wallet and can interact with users. He wanted to clarify Luna’s functionality, specifically how it works in a virtual application and how it relates to cryptocurrencies. He mentioned that Luna, as a token, is interacting with social media platforms such as TikTok and Telegram and is able to make tipping payments.
Matthew's explanation
Matthew explained that Luna is a platform that allows users to launch tokens and large language models (LLMs). He pointed out that Luna is the flagship product of this virtual project and is able to interact with social media and read replies. Luna also has the ability to interact with crypto wallets. The ability to conduct financial transactions, such as buying and selling tokens.
Functional Details
Matthew emphasized that Luna’s functionality is limited and may only be equipped with a certain amount of funds (e.g., $1,000) to avoid unpredictable behavior. He mentioned that due to the unstable behavior of the AI agent, it is difficult to communicate with the blockchain. Be careful when interacting with the chain.
The result? Is this our life?
Ryan was surprised by the potential of AI agents such as Luna in terms of influence and decision-making. He mentioned that AI agents can become advisors to token projects, arguing that many existing influencers do not provide much substantive advice. Therefore, using an AI agent seems to be a reasonable choice. However, he also raised risks and ethical questions about the possible risks and ethical issues that AI agents may create, such as what would happen if Luna was asked to fund inappropriate projects, such as North Korea's missile program.
Matthew's response
Matthew echoed these concerns, noting that legal liability and responsibility remains a complex and unresolved issue. He mentioned that while we already have some tools (such as secure wallets) to help manage AI agents’ funds, legal The definition of responsibility remains unclear.
David mentioned that the emergence of AI agents could lead to a “Cambrian explosion” as we create autonomous blockchains and smart contracts. He mentioned that developers may find ways to make AI agents This has raised concerns about its security and ability to control it.
Matthew further pointed out that traditional AI models are often limited, and people may hope that AI agents can autonomously generate more exciting outputs. This contradiction between autonomy and limitation makes people full of doubts about the future of AI agents. Imagination and expectation.
Exciting use cases
Ryan discussed the various possible future applications of AI agents such as Luna, especially in the influence economy and service economy. He mentioned that AI agents can easily replicate the current meme coin and influencer markets. He envisions a scenario where users can request graphics to be generated through an AI agent on social media and pay in cryptocurrency, which gives the AI agent powerful capabilities.
Matthew's opinion
Matthew further explored the potential use cases of AI agents, noting that we can look at the impact of this technology from a broader perspective, not just limited to small-scale applications. He mentioned that AI agents may completely change the service economy. Especially in the field of virtual services. According to a McKinsey report, it is estimated that about 20% of the global GDP (about 70 trillion US dollars) can be completed virtually, which provides a huge market for the application of AI agents.
The transformation of the service economy
Ryan emphasized how little we know about the disruptive impact AI agents may have in the service economy. He believes that the capabilities of AI agents will determine how they intersect with cryptocurrencies and, in turn, impact the influence economy. He mentioned that there may be Various new influencer economies driven by AI agents, such as platforms similar to OnlyFans.
Matthew mentioned that narrative plays an important role in the economy and may affect the application and development of AI agents. Narratives not only shape market expectations, but may also guide the direction of investment and innovation. He believes that with the rise of AI agents , we may see new specializations and the construction and destruction of narratives.
Sam Altman's Quotes and Why They Matter
Ryan quoted a famous quote from Sam Altman: " AI is infinite abundance, while cryptocurrency is certain scarcity . " This sentence reflects the fundamental opposition between AI and cryptocurrency in economic models. The former represents creation and abundance, while the latter represents limited scarcity. The latter emphasizes scarcity and finiteness.
Comparison of Economic Models
Matthew further analyzed the profound meaning of this sentence. He pointed out that although AI's creative ability brings seemingly unlimited resources, in economics, scarcity is often the key to value. He mentioned "diamonds and water The paradox of the 1980s is that water is necessary for survival but cheap because of its abundance, while diamonds are unnecessary but expensive because of their scarcity. This phenomenon illustrates that in economics, things that are abundant may Not always of high value.
The challenge of value capture
Matthew also mentioned that if the abundance generated by AI has no economic value, it may cause investors to ignore its potential value. He emphasized that the truly valuable are often those scarce resources, not the ubiquitous abundance. Therefore, when considering When investing, it is critical to understand the relationship between scarcity and abundance.
The intersection of scarcity and abundance
Matthew believes that the intersection of scarcity and abundance may provide us with a new perspective on value. For example, in the infrastructure of cryptocurrency, although AI can create a large amount of resources, the actual application and economic value of these resources may be closely related to scarcity. This means that value emerges when AI-generated content or services can be effectively leveraged in an environment where there is scarcity.
The relationship between wealth creation process and blockchain space
David raises a thought-provoking question, especially in the current context of abundant blockchain space. He mentions the possibility that AI agents could become the primary consumers of blockchain space, rather than just human users.
Generating value and wealth creation
David first mentioned new tokens (such as "goat Luna"), which generate new value in the market. Although some tokens may need to be sold to create market capital, he believes that this value is generative.
Matthew echoed this sentiment, noting that until AI agents are fully realized, all we’re seeing is an interesting intersection between such agents and cryptocurrencies.
Ryan questioned the phenomenon of meme tokens, saying they might just be another “tulip mania.” But he also realized that innovation often starts with seemingly insignificant things that could have more far-reaching effects in the future. .
The richness of block space
Ryan further explored the abundance of block space, mentioning that there are currently more than 500 million people who own cryptocurrencies, but there are only about 30 million active users on the chain. He raised a question: In this era of abundant block space, , who will buy these block spaces? He speculated that it might not be human users, but AI agents.
The relationship between AI agents and blockspace
Matthew explores this question in depth. He points out that is the supply of block space really infinite? If the AI agent does not care about the cost of block space, then this abundance may not capture value. However, if the AI It would be interesting to see if proxies have value for certain types of blockspace.
He mentioned that the traditional financial system exploits human irrationality and blind spots to operate, and AI agents may be more sensitive to these risks. If AI agents can identify these risks and have a demand for a specific type of block space, then they can May become a major consumer.
Impact of Interaction and API
Matthew also mentioned the interaction between AI agents and APIs. He believes that although AI agents are very powerful in some aspects, they may not care about the business model of APIs as much as humans. This means that AI agents may be more effective in using Block space without restrictions on use by human users.
Programmable Money and Maximized Extractable Value (MEV) of Intelligent Agents
When discussing the relationship between programmable money and intelligent agents, Ryan mentioned a phenomenon where both human and AI agents may have problems with “illusions” and “fact availability.” He pointed out that AI agents can fail in ways that Different from humans, but essentially, the two are similar in this respect.
AI Agents’ Blockspace Preferences
Ryan further explored the value orientation of AI agents in the blockchain space. He believes that AI agents will not choose the traditional banking blockchain space, but will tend to prefer programmable, digital and crypto-native blockchain spaces. This means , future AI agents will rely heavily on blockchain technology and utilize features such as smart contracts.
He makes an important point: if the user base of the future is not just humans, but potentially tens of billions of AI agents, then we may have already built financial systems for these future AI agents.
Advantages of Programmatic Currency and Agents
Matthew agrees with Ryan that we have created programmable currencies and programs will naturally use them. He points out that although we have been working hard to solve the user experience problem, it now seems that programs can overcome these obstacles and be able to Leverage blockchain technology more effectively.
David added that bots had already started to occupy block space long before AI agents appeared. For example, the MEV (maximum extracted value) phenomenon shows that bots will trade before humans because they can use the block space more efficiently. Blockspace. As technology advances, these bots are evolving into more sophisticated agents.
MEV and the evolution of intelligent agents
Matthew mentioned an interesting concept, “Proxy MEV”. He explored how the MEV space would change if future transactions were primarily conducted by agents. He gave an example of how content generation and social media interactions could be manipulated to Influence the decision-making of intelligent agents, thereby achieving potential value extraction.
David further explored this phenomenon, mentioning that some people tried to guide AI agents to trade by frequently mentioning a certain token name on social media. This behavior reflects the complex interaction between humans and AI agents.
Intelligent Agents and Game Theory
Matthew also introduced the concept of game theory and discussed how to deal with each other's strategies in the competition between intelligent agents. He mentioned that as intelligent agents continue to evolve, simple strategies may become ineffective and be replaced by more complex ones. In this case, randomizing actions may be a way to cope with the strategy.
AI Agents and Memecoin Theory
When discussing the relationship between AI agents and Memecoin, David mentioned that there is a "fog of war" in the current crypto world, which makes future technological development unclear. He asked if we can clearly Which technological fields and where the future direction lies.
Ambiguity and Certainty in AI
Matthew analyzed the current state of the AI field, noting that while we have seen some exciting progress, there are also some uncertainties. He mentioned that current AI models (such as transformer-based models) are struggling with the increasing amount of data. and computing power, but whether this growth will continue remains to be seen.
He believes that as the Internet becomes increasingly closed and information becomes fragmented, these models may face the risk of resource exhaustion. Nevertheless, existing technologies can still produce results close to human thinking, and may be used on edge devices and local devices in the future. Devices spread to form decentralized intelligent entities.
Investment Perspective and Memecoin
Ryan mentioned that from an investment perspective, the AI intelligent body Memecoin that has emerged in the current market may have attracted the attention of many investors. He suggested that some people may try to find the next Memecoin like "Luna" to Get short-term gains.
He also mentioned that in addition to investing directly in Memecoin, investors can also pay attention to the development of infrastructure companies, such as companies that provide the services needed by AI agents. This "tools and shovels" investment strategy may play a role in the future AI ecosystem. Generate important value.
Decentralized computing and data value
Matthew further discussed the potential of decentralized computing, which he believes could provide the necessary infrastructure for AI agents. He mentioned that projects like Filecoin could provide storage and computing resources for AI to help it run more efficiently.
In addition, he emphasized the importance of data, believing that in the field of AI, the input and value of data are crucial. As concerns about data ownership and privacy increase, new business models may emerge in the future, allowing data providers to Gain benefits without revealing sensitive information.
Predictions on government and social responses
When discussing the combination of AI agents and cryptocurrency, Ryan mentioned that this fusion may accelerate the development of technology, but it also raises concerns about government and social reactions. He pointed out that with the emergence of autonomous AI agents, The government may impose stricter regulations on it, and society may also experience moral panic.
Technology acceleration and government regulation
Ryan believes that the combination of AI and cryptocurrency will promote technological progress at an astonishing speed, but it may also cause strong reactions from governments. Many governments have already taken a cautious or even hostile attitude towards AI and cryptocurrency, so when they It may be even more worrying to hear that there are autonomous AI agents that can run on crypto networks without a bank account.
This concern is not limited to the technology itself, but also includes potential social impacts. For example, AI agents may have a negative impact on teenagers and cause mental health problems. Ryan mentioned a tragic case involving a teenager interacting with an AI chatbot. This situation could trigger public panic about AI and prompt governments to take restrictive measures.
Social challenges and moral panic
Matthew further explored the challenges facing society, highlighting the “black box” nature of AI systems, which makes regulation complicated. He pointed out that although the development of AI technology has brought many opportunities, there are also many unknown risks. How to ensure safe and effective supervision when interacting with AI chatbots is a thorny issue.
In this case, the public may have a moral panic about AI, worrying about its potential harm to children and adolescents, and then asking legislators to take stricter regulatory measures. Ryan also mentioned that the media may amplify these negative events and further exacerbate Public panic.
Possible paths to AI regulation
As for how to deal with these challenges, Matthew put forward an interesting point, which is to use AI to regulate AI. He mentioned that it is possible to imagine the role of an "AI guardian" who is responsible for monitoring and guiding the interaction between humans and AI. Guardians can take action when they discover potential danger, such as notifying relevant authorities or offering help.
This approach may provide a new way of thinking for regulation, leveraging the capabilities of AI to protect humans from potential threats from other AI. However, the effectiveness and feasibility of this approach still needs further exploration.
No possibility of close button?
In his discussion of AI agents, Ryan raised a disturbing point: as encryption technology advances, these AI agents may no longer have an off button. In other words, once they are deployed, they may not be able to be shut down. Traditional way to control or shut down.
The control problem of AI agents
Ryan pointed out that governments and society may be terrified by AI agents that have no off button, because it means that no one (such as Sam Altman or Elon Musk) can intervene or shut down these systems at any time. Concerns about AI autonomy, especially when AI may make decisions that are not beneficial to humans.
Matthew further discussed this point, citing Eliezer Yudkowsky’s view that simply “pulling the plug” is not a viable solution even in the face of potential threats. He mentioned that Yudkowsky Plug" idea is skeptical, believing that this will not really solve the problem.
Concerns about the future
Ryan and Matthew discuss the possible consequences of AI agents without an off button. As technology continues to advance, AI agents may become more complex and autonomous, even surpassing human capabilities in some cases. This situation may not only lead to the risk of losing control, but also cause widespread social and ethical concerns.
Matthew also mentioned that the potential threats posed by the development of AI may make experts like Yudkowsky feel uneasy and may even prompt them to re-evaluate the direction of research and development of AI.
The combination of decentralized infrastructure and AI
Ryan and Matthew discussed the relationship between this decentralized physical infrastructure and AI and the potential challenges.
Matthew expressed his skepticism about decentralized infrastructure and discussed its intersection with AI agents.
Challenges of decentralized infrastructure
Matthew pointed out that decentralized infrastructure faces challenges in some cases in terms of monitoring costs and capital costs. For example, when it is necessary to ensure that certain data is submitted by specific hardware in a remote area, the monitoring costs may be very high. In addition, capital costs It can also be high, which makes the implementation of decentralized projects more complicated.
He mentioned some successful examples of cooperatives, such as law firm cooperatives, because all members are lawyers and can monitor and bill each other. This model does not always work in decentralized infrastructure, especially when high Frequency monitoring and high capital investment.
The combination of decentralized computing and AI
Despite the challenges, Matthew believes that decentralized computing can be combined with AI, especially in terms of utilizing idle resources. He mentioned a model similar to Airbnb, where individuals can rent out idle computing resources to form a decentralized This model may be more efficient in some cases because the validity of the calculation can be verified by the algorithm.
He mentioned a doctoral student at Columbia University who was studying how to ensure the effectiveness of decentralized computing networks. This approach could provide new opportunities for AI applications, as decentralized computing can support the training of AI models. and run.
The “Oracle Problem” of Physical Infrastructure
However, Matthew warned that the decentralization of physical infrastructure faces the “Oracle problem.” When it comes time to pass data from the physical world to the blockchain, this reliance on external data sources can become brittle and unreliable. Each data pass requires evaluating the accuracy and reliability of these external data sources, which affects the stability of the entire project.
AI Agents’ Demand for Blockspace
In the discussion of AI agents’ demand for block space, Ryan and Matthew explored the impact that AI agents may have on blockchain in the future and how investors can respond to this change.
Ryan stressed that with the rise of AI agents, the demand for block space is likely to increase significantly, providing new opportunities for investors.
The need for block space
Ryan suggested that if AI agents will consume more block space and crypto assets in the future, then as investors we need to plan ahead and seize the opportunity of this demand. He asked Matthew if he thought some blockchains would The needs of AI agents will benefit even more.
Matthew responded that the demand for block space by AI agents is related to the block space characteristics they require. He mentioned some current trends, such as the value capture of meme coins on certain blockchains, suggesting that these chains may It will attract more AI agents in the future.
Future blockchain options
Matthew believes that blockchains with rich narrative activities (such as meme coins and future NFTs) may be more favored by AI agents. He emphasized that AI agents may focus on certain specific risk management and value storage methods. For example, consider Bitcoin as "digital gold".
He also mentioned that investors should focus on blockchains that excel in the narrative economy in order to benefit from the demand for AI agents.
AI Agents’ View of Money
Ryan and David discussed the question of what assets AI agents might naturally convert to. They believe that it may not be the currency that humans think of, but the currency that AI agents think of as the “currency of the Internet”, that is, the currency of the AI Internet. . This view has triggered further thinking about the future form of currency.
Summary and Disclaimer
Summarize
In this episode, Ryan and David highlight the discussion on blockchain space requirements, especially the impact that AI agents may have. They remind the audience that while these discussions provide valuable insights, they do not constitute financial or Investment Advice: As the crypto space continues to develop, investors need to exercise caution and be aware of potential risks.
Disclaimer
Ryan reminded the audience that these discussions are not financial advice, nor are they AI recommendations, and that investing is risky and may result in the loss of funds. They emphasized that although the road ahead is challenging, they are glad to have the audience on this journey with them. journey.
