Recently, more and more new projects have emerged in the AI ​​track, and attention is not enough. In fact, every wave of popularity is like this. At this time, I have a few past experiences for reference. I will also think and operate in this way. There are only a few. First, the main funds are on the earliest, most focused, and most mature projects. At present, there are three virtual ai16z swarms. The market value of the first two is large, and the market value of the latter two is smaller in turn. The smaller the market value, the more violent the rise will be. This is normal. Among these three, only virtual has a top agent like aixbt. The other two need more time. Second, a wave of big trends in large-scale land. Is this wave over? I think there are some staged signals not long after the beginning. 1. B installed on B 2. up on B 3. cb on B I mean the three giants above. When these signals appear, if there is a short-term surge, you can do a wave of shipments, and don’t care about other times (you can do this yourself, and it is right to ship at any time). In fact, because AI is the main narrative, as long as BTC does not enter a bear market, funds will continue to speculate. So you can do small bands, and for large bands, just look at the big market! 3. The more new B types, the more traps there are. So, for new projects, especially refurbished projects (that is, those who did that before, now the agent is popular, and they jump out to do chains, infrastructure, and pads are basically pitfalls), because like virtual, they have been doing it for a long time to get the current harvest, and the new ones are in a hurry to release them, and their purpose is only one-your usdt. You can participate in new projects, and it is recommended to use only 1% of the opportunities. If you really want to make money, it is better to focus on the targets I mentioned, and buy more when there is a big drop, and buy less when there is a small drop. The three major platforms are virtual, ai16z, swarms, monomers, aixbt, and infrastructure cookies. They are similar. Others, such as games, are actually clusters, etc. They are all good, but they are all included in others. Focus on the above. Understand the logic, know the target, it’s nothing more than how to get on the bus. I take the one I like and get on the bus. Among the three major platforms, swarms pays more attention. Wang Feng himself is a programmer and has a large team. After he gets on the bus, there is no problem in further verifying the code. In addition: I said that the agent can be implemented. In fact, I don’t know whether this implementation is good or bad, but aixbt has already come out. As for the others, I am not particularly concerned, because I estimate that we will stop hyping it up for a few months. When it is really implemented, such as public chains, then it may be time to ship.