Original title: Focus areas at the intersection of crypto and AI
Original author: Kuleen, Head of DePIN at Solana Foundation
Original translation: Yuliya, PANews
Currently, the intersection of AI and crypto technology is entering an 'Cambrian explosion'-style experimental phase. This article from the Solana Foundation elaborates on three key development directions of the AI + crypto fusion.
TLDR
1. Build the most vibrant intelligent agent-driven economy on Solana
Truth Terminal has already proven the feasibility of AI agents operating on-chain. Experiments in this field are continuously pushing the boundaries of agents' operations on-chain, and this field not only has immense potential but also a broad design space. This has become one of the most groundbreaking and explosive directions in the crypto and AI fields, and this is just the beginning.
2. Enhance LLMs' capabilities in Solana code development
Large language models have already shown excellent performance in code writing, and will further improve in the future. With these capabilities, Solana developers are expected to see their efficiency increase by 2-10 times. Recently, establishing high-quality benchmarks to evaluate LLMs' understanding and writing capabilities for Solana code will help understand LLMs' potential impact on the Solana ecosystem. High-quality model fine-tuning schemes will be validated in benchmark tests.
3. Support an open and decentralized AI tech stack
"Open and decentralized AI tech stack" includes the following key elements:
· Training data acquisition
· Training and inference computing capabilities
· Model weight sharing
· Model output verification capability
The importance of this open AI tech stack is reflected in:
· Accelerate model development innovation and experimentation
· Provide alternatives for users who do not trust centralized AI
1. Build the most vibrant intelligent agent-driven economy
Discussions about Truth Terminal and $GOAT have already been extensive, so there's no need to elaborate further. But it is certain that when AI agents begin to participate in on-chain activities, a new world of possibilities has already unfolded (notably, agents have not yet taken direct action on-chain).
While it is not yet possible to accurately predict the future development of agents' on-chain behavior, we can glimpse the broad prospects of this design space by observing the innovations that have already occurred on Solana:
· AI projects like Truth Terminal are developing new types of digital communities through Meme coins like $GOAT.
· Platforms like Holoworld AI, vvaifu.fun, Top Hat AI, Alethea AI allow users to easily create and deploy intelligent agents and their related tokens.
· AI fund managers trained based on the personality traits of well-known crypto investors are emerging, with the rapid rise of ai16z on the daos.fun platform creating a new ecosystem of AI funds and agent supporters.
· Furthermore, gaming platforms like Colony allow players to participate in the game by guiding agent actions, often resulting in unexpected innovative gameplay.
Future development directions
In the future, intelligent agents could manage complex projects that require multi-party economic coordination. For example, in the field of scientific research, agents could be responsible for finding therapeutic compounds for specific diseases. Specifically:
· Token fundraising through the Pump Science platform
· Utilize raised funds to pay for access to paid research materials, and computing costs for compound simulations on decentralized computing networks like kuzo, Render Network, io.net, etc.
· Recruiting humans to perform experimental verification work (e.g., running experiments to validate/build simulation results) through bounty platforms like Gib.Work
Apart from complex projects, agents can also perform simple tasks such as building personal websites, creating artworks (like zerebro), with limitless application possibilities.
Why does it make more sense for agents to perform financial activities on-chain than through traditional channels?
Agents can fully utilize both traditional financial channels and cryptocurrency systems. However, cryptocurrencies have unique advantages in certain areas:
· Micro-payment applications — Solana excels in this regard, with applications like Drip already proving this.
· Speed advantage — instant settlement features contribute to agents achieving maximum capital efficiency
· Entering capital markets through DeFi — this may be the strongest reason for agents to participate in the crypto economy. The advantages of cryptocurrencies become even more evident when agents need to engage in financial activities beyond payments. Agents can seamlessly mint assets, trade, invest, engage in lending operations, use leverage, and more. Particularly, Solana is well-suited to support these capital market activities due to its existing top-tier DeFi infrastructure on its mainnet.
From the perspective of technological development laws, path dependence plays a key role. Whether a product is optimal is not the most important; what matters is who can first reach critical mass and become the default choice. As more and more agents earn through cryptocurrencies, crypto connections are likely to become a core capability of agents.
The foundation hopes to see
The Solana Foundation hopes to see agents equipped with crypto wallets conduct bold innovative experiments on-chain. The foundation does not overly limit specific directions here, as the possibilities are indeed too broad — believing that the most interesting and valuable agent application scenarios are likely to be unforeseen at this moment.
However, the foundation is particularly focused on exploring the following directions:
1. Risk control mechanisms
· Although current models perform well, they are still far from perfect.
· Cannot grant agents completely unrestricted freedom of action
2. Promote non-speculative use cases
· Purchase tickets through xpticket
· Optimize stablecoin portfolio yields
· Order food on DoorDash
3. Development progress requirements
· At least reach the prototype stage of the testnet
· Preferably already running on the mainnet
2. Enhance LLMs' ability to write Solana code, empowering Solana developers.
LLMs have demonstrated strong capabilities and are making rapid progress. In the application field of LLMs, the area of code writing may see a particularly steep progress curve, as it is a task that can be objectively assessed. As mentioned below, "Programming has a particularly unique advantage: the potential for superhuman data scaling through 'self-play.' Models can write code and run it, or write code, write tests, and then check for self-consistency."
Today, although LLMs are still not perfect in code writing, with obvious shortcomings (for example, performing poorly in bug detection), AI-native code editors like Github Copilot and Cursor have fundamentally changed software development (even the way companies recruit talent). Given the expected rapid progress rate, these models are likely to completely transform software development. The foundation hopes to leverage this advancement to enhance Solana developers' work efficiency by an order of magnitude.
However, several challenges currently hinder LLMs from achieving excellence in understanding Solana:
· Lack of quality raw training data
· Insufficient number of verified builds
· Lack of high information value interactions on platforms like Stack Overflow
· Historically, Solana's infrastructure has developed rapidly, meaning that even code written 6 months ago may not fully meet today's needs.
· Lack of methods to assess the model's understanding of Solana
The foundation hopes to see
· Help acquire better Solana data on the internet
· More teams releasing verified builds
· More people in the ecosystem actively asking good questions and providing high-quality answers on Stack Exchange
· Create high-quality benchmarks for evaluating LLMs' understanding of Solana (RFP to be released soon)
· Creating LLM fine-tuning models that perform well in the above benchmarks is crucial, and more importantly, accelerating Solana developers' work efficiency. Once high-quality benchmark tests are established, the foundation may offer rewards to the first model that reaches the benchmark testing threshold score.
The ultimate significant achievement will be: a completely AI-created, high-quality, differentiated Solana validator node client.
3. Support an open and decentralized AI tech stack
In the field of AI, the long-term power balance between open-source and closed-source models remains unclear. There are indeed arguments that support closed-source entities will continue to stay at the technological forefront and capture the main value of foundational models. The simplest expectation now is to maintain the status quo - tech giants like OpenAI and Anthropic drive frontier development, while open-source models quickly follow and gain unique advantages through fine-tuning in specific application scenarios.
The foundation is committed to closely integrating Solana with the open-source AI ecosystem. Specifically, this means supporting access to the following elements:
· Training data
· Training and inference computing power
· Model weights
· Model output verification capability
The importance of this strategy is reflected in:
1. Open-source models accelerate innovation and iteration
The rapid improvements and fine-tuning of open-source models like Llama by the open-source community demonstrate how the community can effectively complement the work of large AI companies, pushing the boundaries of AI capabilities (even Google researchers pointed out last year that 'we have no moat regarding open-source, and neither does OpenAI'). The foundation believes that a thriving open-source AI tech stack is essential for accelerating progress in this field.
2. Provide choices for users who do not trust centralized AI
AI may be the most powerful tool in the arsenal of dictatorships or authoritarian regimes. State-sanctioned models provide an officially sanctioned 'truth,' serving as an important control vehicle. Highly authoritarian regimes may have superior models because they are willing to ignore citizen privacy to train AI. The use of AI for control is an inevitable trend, and the foundation hopes to prepare for this by fully supporting an open-source AI tech stack.
Several projects in the Solana ecosystem are already supporting the open AI tech stack:
· Data collection — Grass and Synesis One are advancing data collection
· Decentralized computing power — kuzo, Render Network, io.net, Bless Network, Nosana, etc.
· Decentralized training frameworks — Nous Research, Prime Intellect
The foundation looks forward to seeing
Hoping to build more products at all levels of the open-source AI tech stack:
· Decentralized data collection: For example, Grass, Datahive, Synesis One
· On-chain identity: Protocols that support wallet verification of human identity, verifying AI API responses, enabling users to confirm they are interacting with LLMs
· Decentralized training: Projects like EXO Labs, Nous Research, and Prime Intellect
· IP infrastructure: Allowing AI to license (and pay for) the content it uses
Original link
