Author: DeFi Teddy
The valuation of AI Agents has been a roller coaster, where is the future path? As a Web3 AI entrepreneur and investor, here are some personal thoughts:
1. Why the crash
The crash of AI Agents has two triggering factors, one is that the trump meme coin has drained the liquidity of the market, and on the day of trump's launch, many people sold their AI coins to go 'All In' on trump; another more important factor is the launch of low-cost 'deepseek', which has reduced the dependence of AI on computing power, causing the collapse of Nvidia, and the sentiment has also been transmitted to the stock prices of other AI companies and the coin prices of Web3 AI Agents, which is even a blow to the technology innovation led by the United States, and the crypto market and the US (US stock market) are strongly correlated.
However, in the long run, the reduction in the underlying computing power cost is very beneficial for building AI Agents, for example, now the development of AI applications, due to the cost issue of chatgpt, developers can only reduce the number of calls, with 'deepseek', the cost issue is greatly alleviated.
2. The valuation logic of AI Agents
The current market capitalization valuation of AI Agents = (fundamentals + narrative) * market sentiment.
The fundamentals are the current situation of the project, such as cash flow, active user numbers, user usage time, Twitter heat, etc., and the narrative is the imagination of the project's future, if the current fundamentals are general, but the narrative has prospects, the market will also buy in, this is the same as the reason why users buy stocks that are losing money but have development prospects. Market sentiment fluctuates greatly, the market was in a serious FOMO mood recently, so it gave a lot of premium to AI projects; now the FUD is very serious, so the valuation has shrunk significantly.
3. Successful applications of AI Agents
Why focus on the successful applications of AI Agents, rather than the AI Infra? Because with successful applications, the Infra must have value; if we only look at the infra, without successful applications, the business model is not viable.
Currently, @AIxbt_Agent has already proven the product market fit of AI Agents, by quickly building a crypto KOL that is better than humans through LLM, and AIxbt_Agent is in the top 3 in kAIto's mindshare.
In addition, @_kAItoAI, although not an AI Agent, is also a successful application of web3 AI. Previously, kAIto's crypto data analysis did not have a particularly good breakthrough point, but through 'yaps', it quickly found the product market fit, and through LLM technology, it found high-quality KOLs and content, and can also help projects with marketing, some top projects like @StoryProtocol @monad_xyz have established cooperation with kAIto. Allowing both the project side and the users to 'yaps', the ultimate beneficiary is kAIto.
4. Problems faced by AI Agents
The first problem is the market sentiment problem, which can be fixed over time; the second is the liquidity problem, if the overall liquidity of the crypto market has problems, the AI Agents rooted in the crypto market will certainly be affected; the third is that the AI Agents launched by 'pumpfun' have many project plates, without long-term construction plans, squeezing the space of the real long-term builders, which has a certain 'bad money driving out good money' effect; the fourth is that AI's language ability is very strong, but in terms of real trading decision-making, it does not have outstanding capabilities, and may even have delusions that lead to investment errors.
5. Future outlook
In the short term, AI Agents may be subject to the above problems, but in the long run, they have a very large development space, which depends on the progress of AI technology itself and the time and cost of training in the web3 domain. @AIxbt_Agent has already become a successful crypto KOL, the next stage is DeFAI, but it is still in the early stage. Recalling that DeFi was first proposed in the summer of 2018, and truly flourished two years later in the DeFi Summer, perhaps DeFAI also needs such a cycle.
