jakiro
Source: Pantera Capital October Blockchain Letter; Compilation: 0xjs@来 来 来 来 来
Crypto: AI Gold Gold Herizes
Author: Matt Stephenson, Pantera Capital Research Partner; Ally Zach, Pantera Capital Research Engineers
“AI is infinitely rich, and Crypto is absolutely scarce.”
Sam Altman’s observation in 2021 has since become the mantra of these two technical enthusiasts.At first glance, richness seems to be more influential than compulsory scarcity, which indicates that AI may be more cautious investment.In fact, Nvidia’s market value is larger than the entire cryptocurrency.
But Altman’s remarks are reminiscent of Adam Smith’s “diamond and water paradox”.Smith pointed out that although water is important for survival, the rich water resources make the water almost worthless.
On the contrary, although diamonds do not have much practical use, they are very valuable due to scarcity.This paradox shows,Even if AI becomes as important as water, its market value may still be limited.In contrast, the scarcity of cryptocurrencies is more strategic and valuable than the original.
Large -scale language models (LLM) have achieved remarkable achievements, including Turing test, which is reported that it is better than humans in standard IQ tests.But this raises a question: Can they distinguish between human and intelligent AI (in Turing test), can they distinguish between intelligent AI?If humans cannot distinguish, the improvement of AI performance in the future may have a decreasing return in the income that consumers can perceive.
As the TV resolution jump from 4K to 8K for ordinary audiences, the improvement of the visible improvement is minimal. The difference between the high -performance AI model and the slightly advanced model may be difficult to detect for most users.This may lead to most AI markets. The most advanced model will only be used in professional applications in the field of research, industry or government, and more cost -effective “good enough” model will become the standard for daily use.Top AI models may become “expensive boutiques for mainstream consumers will never consider upgrading.”
Therefore, even if we speculate that the potential growth of AI, we should consider another choice: the powerful features known to the current AI have already existed and become increasingly commercialized.This is where encryption and artificial intelligence (“Crypto X AI”) really became the focus.Crypto’s potential may not bet on the AI MEME value, but the practical value acquisition mechanism for AI distributed future.Once everyone has a 4K TV at home, its value lies in what we use with them.
By acting as an important and reliable input of AI, and the orbit of distributed AI coordination and transactions, cryptocurrencies are closer to conservative “Iron Holy Ho” bet on AI.This may surprise those investors who treat Crypto X AI as a potential growth of AI.But it is interesting that in the past six months, I think Nivine as an agent of AI growth, cryptocurrencies look more like hedging AI growth emotions, rather than high -Beita -value investment.
We will first evaluate the bright prospects of “AI Smart” and how encryption technology will play a role.We will then discuss the potential of encryption technology to support the current input of AI: data, computing and models.
AI Smart: Using programmable currency procedures
Author: Matt Stephenson, Pantera Capital Research Partner
Last year, before most people talked about the AI intelligence on the blockchain, I collaborated with others to write a paper, which was accepted by Neurips, the top AI meeting in the United States.Since then, I have been fortunate to have participated in encryption and smart AI activities at Stanford University, Columbia University, Cornell University, and Berkeley University, and delivered a speech. In addition, I also participated in many technical and investment meetings.Next week, I will give a speech with Professor of Oxford University, Chairman of IEEE and GBBC. All these speeches will be to better understand, explore and exchange what the future of smart AI, and how it intersects with blockchain.Of course, I also invest in this future, including investment in smart infrastructure such as SENTIENT and other uninteresting positions.
The future has come.Although OpenAI said that AI intelligence can not be ready until 2025, but in the field of cryptocurrencies, we now have AI smart parties that trades and explore in blockchain space.A AI intelligent body that promotes its tokens (Note: Truth Terminal) currently has about $ 300,000. When you read this article, you read this article, andIt may become the first AI smart man.
But what are these intelligences?How are they different from the “robots” we are more familiar with?
Agent is not just a robot
Definition “Agent” is more delicate than it looks.The definition of Agent in the field of artificial intelligence is not practical: “Any things that perceive the environment through sensors and take action through the actuator.” Economists’ views on Agent are closer to what we want: “Agent is a person who takes action in a specific decision -making field.“
If Agent represents you, then the robot is essentially an agent that is difficult to communicate.First of all, you must write code for robotic execution, which means to communicate with (programming) language that most people do not understand.For those who understand language, they still have to write procedures for robots to do under various conditions, which means that these conditions must be specified in advance.Both of them are communication costs.
For example, if you have a friend to go abroad, you ask him to help you buy a souvenir.If your friend is like a robot, he will ask you to write a program, specifying what souvenirs they should buy for you.If your friend is like an agent, what will happen?Then you can use language to ask, you can believe that your friends will buy you what you want.Using languages, there is no need to explain the preferences you may receive abroad, which can reduce communication costs.Obviously, this is a better agent.
You must understand the conditions of the robot as an agent in advance (because you must program them programming) in advance.Then, the fact that it is necessary to programming the robot means that it is out of reach for those who are not programmed.We will turn to AI Agent modeling to reduce the cost of communication and release the corresponding economic value.
Although the communication costs of existing robots are very high, more than 2 trillion dollars of cryptocurrency stable currency transactions have seem to be robot transactions.As the robot becomes a better agent, maybe you can trade USDC and USDT according to relative risk, and we should expect this number to increase.
AI Smart will use encryption technology
One of the reasons for AI intelligence to help cryptocurrencies is that it helps to alleviate the problem of encrypted currencies’ infamous user experience problems.The complexity of blockchain interaction, wallet management and decentralized financial protocols has been widely used for a long time.AI Smart can act as an intuitive interface and transform user intentions into precise technical operations required for the blockchain.They can guide users to complete complex transactions, explain risks, and even recommend the best strategies according to market conditions and user preferences.
Another reason is that the smart body cannot have a bank account, but can be traded with wallets.This restriction of traditional financial systems is completely in line with the spirit of cryptocurrencies.In the encrypted world, intelligence does not need to obtain a permit from the central institution to operate.They can interact directly with smart contracts and decentralized agreements to represent users hold and manage digital assets.This has opened up new possibilities for automated wealth management, all -weather transactions and personalized financial services that fully operate in the encryption ecosystem.
Finally, the mature intelligent ecosystem means that smart parties need to transaction and coordinate with each other.As a programmable and online international legal system, modern intelligent contracts are very suitable for this task.AI intelligence can use encrypted infrastructure to participate in complex multi -party transactions and protocols.They can negotiate clauses, execute transactions, and even resolve disputes within the parameter range set by the human commission.This creates a new paradigm of independent economic activities. Agent can establish temporary alliances, collect resources, and cooperate with the tasks that humans cannot or cannot manage directly.
We believe these activities will appreciate the encryption infrastructure.But there are also indirect effects, making the encryption itself better.For example, due to the attention restrictions in encryption, decentralized autonomous organizations (DAOs) have been in an unconventional state.DAO (each smart body represents the interests of DAO voters), which is actively managed by AI Smart Networks), will change the rules of the game.These intelligences can analyze proposals, distribute resources and execution strategies at the speed and scale beyond human capacity, while abide by the core principles and goals of their human creators.
AI Smart and cryptocurrencies are not just a perfect combination, they are two technologies that need to be mutually needed.Intelligence requires programmable currency to operate independently in the digital economy.Cryptocurrencies need AI to improve the user experience and fulfill their commitment to everyone bring the financial revolution.With the development of this synergy, we may see that core blockchain infrastructure such as Solana, Ethereum, Near, and Arbitrum has become the main beneficiary of this new smart body -driven economy.They are going to achieve this goal by promoting decentralized applications that promote smart transactions, hosting agents to interact with them, and provide the security and transparent environment required for smart body coordination.With the increase of smart activities, these networks may see increased transaction volume, increased demand for native currencies, and enhanced network effects.This is not only related to technical compatibility -but also to create a new economic paradigm. Among them, AI and cryptocurrencies work together to make finance more efficient, easier to acquire, and may even be a bit sci -fi.
Crypto technology helps the current AI
Author: Ally Zach, Pantera Capital Research Engineer
Imagine that you are about to get a major breakthrough, but you find that the tools you need are out of reach.Innovation often gives people this feeling -a journey full of breakthrough climax and challenging the trough.Taking the automotive industry as an example, the exploration of more efficient engines was once in a dead end.Engineers are eager to break through the limits, but the required materials have not yet existed.The progress has stagnated until the new alloy and composite materials have re -stimulated the innovation engine.Similarly, new technologies such as encryption may release the potential of AI that has not yet developed.
Over the years, the development of AI has been progressing step by step, first developing slowly, and then rapid development, similar to the S curve.In 2017, we have achieved a key breakthrough. The architecture based on Transformer came into being, just like an influential paper “Attention is everything you needOverview in “.These Transformer completely changed the sequential data processing in the model and achieved efficient training for large data sets.This has triggered the rapid development of strong new LLM and generating AI models.
Although the development of AI has made progress, in order to achieve the next leap, it must overcome major bottlenecks in data, computing and model generation.Combining AI with blockchain technology can help decentralize resources and realize visits to democratization, and open innovation to global contributors.
data
Data is the lifeblood of AI, which is a fuel that promotes its accuracy and reliability.High -quality and representative data are critical to build effective models, but due to privacy issues, limited access and inherent bias, obtaining these data is challenging.In addition, users are increasingly unwilling to share personal information, which makes data collection consume a lot of resources and is often hindered by trust.
Blockchain technology provides a promising solution by introducing decentralized, secure and transparent data aggregation methods.Platforms like Sahara are in line with our long -term strategy to promote AI decentralized infrastructure. It enables individuals to contribute data and monetize it while retaining control.In addition, the tokens to inspire high -quality contributions by rewarding users accordingly.This method helps solve privacy problems by allowing users to own and control their data.It makes data access to democratize and enables small enterprises that previously lacks resources with large technology companies to compete.By motivating data sharing in a secure way, data based on blockchain turns data into a product, enriching the available data pool, and may produce a more stable and fair artificial intelligence model.
However, despite innovation, data aggregation based on blockchain is not an independent solution developed by AI.If it is used alone, actual challenges such as scalability, data quality assurance and integrated complexity will limit its effectiveness.With a huge data set and mature infrastructure, large technology companies still have a significant advantage that it is difficult to compete for decentralized platforms.
Therefore, solutions, including blockchain -based solutions, introduced new data collection and collaboration channels, which are supplements to traditional methods rather than alternatives.The coordinated role between decentralized efforts and mature technical leaders can contribute to the partnership relationship, make full use of the advantages of both parties, and promote the innovation and inclusiveness of AI development.
calculate
The rise and scarcity of GPU costs bring major obstacles to small enterprises in AI development.Due to the strong demand and the supply chain problem, since the outbreak of the epidemic, the price of GPU has continued to rise, and large enterprises have more and more monopolized the use of basic hardware.This limits innovation because many startups and researchers need help to afford the tools for high -end model training.This reduces the diversity of AI research and slows the progress of small institutions.
However, Crypto is likely to create a fair competitive environment by commercializing computing power.Platforms such as EXO and IO.NET are realizing the democratization of GPU access through a decentralized market, and anyone can access or borrow computing resources.Individuals with idle computing capabilities can provide it on the Internet to get returns.High -performance computing commercialization enables more extensive innovators to participate in AI development, breaking the obstacles to restricting access to high -level tools.
In the future, with the increase of the GPU supply, the decentralized computing market may directly compete with traditional cloud services.These platforms have reduced the threshold for access and provided economic and efficient alternatives, enabling people to participate in the AI ecosystem more widely.However, it is still a challenge to ensure that users get reliable computing power.Verifying the GPU standards and maintaining consistent and safe resources is critical to establish trust and prevent fraud.Although decentralized solutions may not replace traditional services, they can provide competitive alternatives because flexibility and cost are more important than guarantee performance.
Model
Today, AI development is usually concentrated in small parts such as OpenAI, Google and Facebook.This concentration limits the opportunity of global innovators and has aroused concerns about whether AI can reflect diverse human values.Concentrated control may lead to a narrow view of the model, and ignore the needs and viewpoints of a wider range of user groups.
The power of allocating AI development through decentralized platforms is changing.SenTient and Near platforms are consistent with our vision of more and more AI on the orbit. They are achieving the democratization of development by creating open source and community -driven ecosystems.Using blockchain technology, they transparently manage contributions to ensure that developers are recognized and compensated through tokens.This allows anyone to build, collaborate, owns and monetize AI products, and start a new era of AI entrepreneurship.The co -author and Near co -founder of NEAR, Illia Polosukhin, the pioneering papers “Attention is the need you need”.Cooperation measures like this aim to combine AI development with extensive human values.
These platforms act as catalysts of change, which promotes both competitive and collaborative AI economies.By expanding participation, they encourage various ideas to flourish, so as to bring more innovative solutions and reduce prejudice in AI models.
Crypto X AI provides unique opportunities for the democratization of AI development, but also brings major challenges.Balanced large -scale collaboration and high -quality, expert -driven work needs are essential to ensure that the model is stable and moral.Through decentralized data access, computing power and model development, the encryption has broken the traditional obstacles, enabling people from all over the world to participate in the development of AI.The influx of this diversified view has promoted collaboration and established a more inclusive ecosystem.Hugging this collaboration model can not only accelerate innovation, but also ensure the future of the global community to create AI.
