jakiro
The Meme market seems to be facing a collapse, and the market’s anxiety begins to spread again.What is the reason?There is no new narrative, but the funds are smart and they all go to blue chip NFTs.
Regarding the new anxiety, I saw several KOLs discussing AI on Twitter today. They believe that Crypto+AI may be the next narrative direction and also the hope for the future.As someone who is engaged in related business, I will talk to you about this topic today.Let me first talk about two of my own experiences:
1. Last year, ChatGPT attracted widespread attention, and then a game called “Stanford Town” was launched on the market.In this virtual town Smallville, researchers used GPT-4 to drive 25 AI agents that live, work, and socialize in the town and show complex behavior patterns such as handling emergencies and waitingUse public facilities and communicate with others.Although the study was not open source at the time, the source code is now open.My friends and I also deployed the game locally and connected to GPT.After playing for less than five minutes, I found that I had paid OpenAI more than ten yuan in fees. After repeated inspections and confirmations, we concluded that the cost of using AI is really not low.
2. This year, we transformed an AI computing power factory in a hydropower station in Yunnan.This was once a Bitcoin mine, but it was abandoned in 2018 for various reasons.This year we re-engineered it and enabled it, using Meta’s Llama3, our customers are big data companies, providing them with computing power support.Although it has its own computing power, the cost of use is still not low. It is a summer flood season. When the power supply is insufficient after October, we have to negotiate to purchase electricity with the local power grid, and the cost will behigher.
Through the above two cases, I want to tell you that the cost of using AI is indeed very high.Of course, some friends will refute that it is not expensive or even free to use ChatGPT chat, using AI tools such as literary pictures, pictures and videos, and creating songs. Of course, these businesses are not expensive, but they can only be limited to these simple businesses. For complex businesses, they are not expensive.The cost is still a big problem, which is why we saw a large number of so-called AI projects that cover GPTs on the market last year. After all, they can’t afford to develop them because they are expensive, and it is impossible for anyone to use them.
The combination of Crypto+AI certainly has the potential to create many innovative products, but the biggest problem at present is that the cost of AI calls is too high, limiting the possibility of large-scale applications.I think both VC and project parties themselves should be clear about this.
Regarding the AI+Crypto track, I think it may take until the next cycle or even later to truly achieve large-scale implementation.Before this, we need to solve two key issues: controlled nuclear fusion (low-cost electricity) and Asic chips specifically targeted at AI (high efficiency and low energy consumption).This will be the basis for large-scale application of AI+Crypto.
At the end of the article, I posted an easter egg, which may contain huge opportunities.
Controllable nuclear fusion may be a future that people in our currency circle cannot reach, and there is basically no chance.But in terms of Asic chips, we have a chance to see breakthroughs.At present, the one in China that has done the strongest in the research and development of AI-specific Asic chips should be Bitmain.That’s right, it’s the Bitmain that makes mining machines, and they have been working on them for many years. Isn’t it a bit unexpected and a little surprised?If there really is one day, you can buy some of them and put them in our computing power factory for hosting, just like digging big cakes before.
