The truth about AI agents: Why is the GOAT, valued at $1 billion, still a mechanical text generator?

Data access is key. Author: MORBID-19 Compiled by: TechFlow Hello everyone, it's a new day and another speculative bet. Recently, AI agents have become a hot topic of discussion. Especially aixbt, this product has been receiving a lot of attention lately. But in my view, this hype is completely meaningless. Let me explain the Bit terminology for those who are not familiar. Once users bridge their assets to the so-called "Bit L2", it is impossible to achieve true "Non-custodial Lending". All "Bit Bridges" or "Interoperability/Scaling Layers" introduce new trust assumptions, with few exceptions like the Lightning Network. So when someone claims that Bit L2 is "Trustless", you can basically assume that this is not true. This is also why most new L2s emphasize that they are "Trust-minimized". Although I am not familiar with Side Protocol, I can almost certainly say that the "Non-custodial Lending" claim made by aixbt is untrue, and this judgment will be correct in 99% of cases. However, I don't completely blame aixbt. It is just acting on instructions: scraping data from the internet and generating seemingly useful tweets. The problem is that aixbt does not truly understand what it is saying. It cannot judge the truthfulness of information, cannot verify its assumptions with experts, and cannot question its own logic or reasoning. Large language models (LLMs) are essentially word predictors. They do not understand the content they output, but rather choose seemingly correct words based on probability. If I wrote an article in the Encyclopedia Britannica about "Hitler conquering ancient Greece and giving birth to Hellenistic civilization", this would become "fact" and "history" for the LLM. Many of the AI agents we see on Twitter are simply word predictors with cool avatars. Yet, the market valuations of these AI agents are skyrocketing. GOAT has reached a valuation of $1 billion, and aixbt's valuation is around $200 million. Are these valuations justified? No one can say for sure, but ironically, I am satisfied with the assets I hold. Data access is key I have always been very interested in the combination of AI and cryptocurrencies. Recently, Vana has caught my attention because it is trying to solve the "Data Wall" problem. The problem is not a lack of data, but how to access high-quality data. For example, would you share your trading strategies for low-liquidity small-cap tokens in public? Would you freely publish the high-value information that usually requires payment? Would you publicly share the most private details of your personal life? Clearly not. Unless your private data can be protected at a reasonable price, you will never easily share these "private data" with anyone. However, if we want AI to reach a level of intelligence close to humans, this data is the most critical element. After all, the core traits of humans are their thoughts, inner monologues, and most secret contemplations. But even accessing some "semi-public" data faces significant challenges. For example, to extract useful data from videos, you first need to generate subtitles and accurately understand the context of the video, so that AI can understand the content. Similarly, many websites require users to log in to view content, such as Instagram and Facebook. This design is common in many social networks. In summary, the main limitations currently faced by AI development include: 1. Inability to access private data 2. Inability to access data behind paywalls 3. Inability to access data from closed platforms Vana provides a possible solution. They overcome these limitations by protecting privacy and aggregating specific data sets into a decentralized mechanism called DataDAOs. DataDAOs are a decentralized data marketplace that works as follows: - Data contributors: Users can submit their data to DataDAOs and receive governance rights and rewards. - Data verification: Data is verified on the Satya network, a network of secure computing nodes that can ensure data quality and integrity. - Data consumers: Verified data sets can be used by consumers for AI training or other applications. - Incentive mechanism: DataDAOs encourage users to contribute high-quality data through transparent mechanisms that manage data usage and training. If you want to learn more, you can click here to read more content. I hope one day aixbt can break free from its "stupidity". Perhaps we can create a dedicated DataDAO for aixbt. Although I am not an expert in the AI field, I firmly believe that the next major breakthrough in AI development will depend on the quality of the data used to train the models. Only AI agents trained on high-quality data can truly realize their potential. I look forward to that moment, and hope it won't be too far away.