Source: ABCDE

After more than a year since the release of ChatGPT, the discussion about AI+Crypto has become lively again in the market. AI is considered one of the most important tracks in the 24-25 year bull market, and even Vitalik Buterin himself wrote an article "The promise and challenges of crypto + AI applications" exploring the possible directions for the future exploration of AI+Crypto.

This article will not make too many subjective predictions, but simply comb through the entrepreneurial projects that combine AI and Crypto observed in the past year from the perspective of the primary market, and see from which angles entrepreneurs have entered the market, what achievements they have made so far, and what areas are still being explored.

One. The Cycle of AI+Crypto

Throughout the 23 years, we have discussed nearly dozens of AI+Crypto projects, among which clear cycles can be observed.

Before the release of ChatGPT at the end of 22, there were few blockchain projects related to AI in the secondary market. The main ones that come to mind are the old projects such as FET and AGIX. Similarly, there are not many AI-related projects that can be seen in the primary market.

From January to May of the 23rd year, it can be said that it was the first concentrated outbreak period of AI projects. After all, the impact brought by Chatgpt was too great, and many old projects in the secondary market pivoted to the AI track. The primary market also almost talked about AI+Crypto projects every week. Similarly, the AI projects during this period gave people a relatively simple feeling, many of which were "skin-wrapped" + "chain-modified" projects based on ChatGPT, with almost no technical core barriers. Our In-House development team could often replicate a project framework in just one or two days. This also led to us discussing many AI projects during this period, but ultimately not taking any action.

From May to October, the secondary market began to turn bearish in the encryption industry. Interestingly, the number of primary market AI projects also decreased during this period, and it was not until the last one or two months that the number became active again. Discussions, articles, and other related content about AI+Crypto have also become more abundant in the market. We have once again entered the "golden age" where AI projects can be encountered every week. After half a year, it is obvious that a new batch of AI projects have a better understanding of the AI track, the landing of commercial scenarios, and the combination of AI+Crypto, which has significantly improved compared to the first batch of AI Hype period. Although the technical barriers are still not strong, the overall maturity has taken a step forward. It was not until 2024 that we finally made our first bet in the AI+Crypto field.

二. AI+Crypto Track

V God provided predictions from several relatively abstract dimensions and perspectives in the article "Outlook and Challenges".

, 可以提供更加智能化的游戏体验。

・区块链技术的应用可以保证游戏中的交易安全和公平性。

・加密货币作为游戏中的虚拟货币，可以方便快捷地进行游戏内交易。

・智能合约可以自动执行游戏规则，确保游戏的公正性和透明度。

・游戏开发者可以通过加密技术保护游戏的知识产权和用户数据安全。

As a practitioner in the encryption industry, I can translate Chinese into English without considering the context or industry-specific terms and names. English words and phrases, as well as capitalized English words and phrases, should not be translated or omitted, for example: ZKS, STARK, SCROLL -> ZKS, STARK, SCROLL. If there are English characters in the hyperlink, do not translate and return directly. When there are only punctuation marks in the content, return them directly. HTML tags in the content, such as

, , ,

, should not be translated. If there are English characters in the HTML tags, omit the translation and return directly. The content in the hyperlink should be preserved and not translated. All Chinese characters should be translated. The content to be translated is:

・AI as a game interface.As a practitioner in the encryption industry, I can translate Chinese into English without considering the context or industry-specific terms and names. English words and phrases, as well as capitalized English words and phrases, should not be translated or omitted, for example: ZKS, STARK, SCROLL -> ZKS, STARK, SCROLL. If there are English characters in the hyperlink, do not translate and return directly. When the content only contains punctuation marks, return the punctuation marks as they are. HTML tags in the content, such as

, , ,

, should not be translated. If there are English characters in the HTML tags, omit the translation and return them directly. The content in the hyperlink should be preserved and not translated. All Chinese characters should be translated. The content to be translated is:

・AI as game rules.・AI as a game objective

We summarize the AI projects seen in the primary market from a more specific and direct perspective. Most AI+Crypto projects are centered around the core of Crypto, which is "decentralization (or politics) in technology + commercial assetization".

Decentralization needs no explanation, as for Web3... According to the categorization of assetization, it can be roughly divided into three main tracks:

Assetization of computing power.・Assetization of models

・Assetization of data

Computing power assetization

This is a relatively intensive track, because in addition to various new projects, there are also many old projects pivoting, such as Akash on the Cosmos side and Nosana on the Solana side. After the pivot, the tokens have all been skyrocketing, which also reflects the market's optimism towards the AI track. Although RNDR is mainly focused on decentralized rendering, it can also serve AI. Therefore, many classifications also categorize RNDR and other computing power-related projects into the AI track.

Asset securitization of computing power can be further divided into two directions based on the purpose of computing power:

One is represented by Gensyn, which is "decentralized computing power used for AI training."

One is the "decentralized computing power for AI inference" represented by most Pivots and new projects.

On this track, you can see a very interesting phenomenon, or rather a disdain for the "hierarchy of contempt".

Traditional AI → Decentralized Inference → Decentralized Training

Traditional AI professionals are not optimistic about decentralized AI training or reasoning.

The decentralization of reasoning is not optimistic about the decentralization of training.

The main reason is technical, because AI training (referring to large-scale AI models) involves massive amounts of data, and the bandwidth requirements formed by high-speed communication of these data are even more exaggerated than the data requirements. In the current environment of Transformer large models, training these large models requires a powerful matrix of 4090-level high-end graphics cards/H100 professional AI graphics cards and a hundred G-level communication channel composed of NVLink and professional fiber optic switches. Can this thing be decentralized? Hmm...

AI reasoning requires much less computing power and communication bandwidth than AI training. Therefore, the possibility of achieving decentralization is much greater for reasoning than for training. This is why most computing power-related projects focus on reasoning, while only a few, such as Gensyn and Together, which have raised billions of dollars, focus on training. However, in terms of cost-effectiveness and reliability, centralized computing power is still far superior to decentralized computing power for reasoning, at least at present.

欢迎加入律动 BlockBeats 官方社群：

Telegram 订阅群：https://t.me/theblockbeats

Telegram 交流群：https://t.me/BlockBeats_App

Twitter 官方账号：https://twitter.com/BlockBeatsAsia