BlockBeats News, December 18th, Binance founder CZ stated in an end-of-year Q&A event:
「AI will be widely used in trading. However, I think there may be several different development paths, which are somewhat different from the prediction market.
Almost all mature traders, large trading firms, and hedge funds have teams managing trading. I expect that they have been using some form of AI, even if unknowingly, possibly in areas such as data analysis. True players will train their AI algorithms, or at least attempt to use AI trading platforms.
AI trading platforms will face a potential contradiction: If you have an AI algorithm that can make a lot of money, why would you sell it as a service to others? Why not trade yourself? The only reason may be that you lack capital. But today, for a good team, fundraising is relatively easy, especially for a truly profitable AI.
If you sell AI as a monthly subscription service to others, it means that selling AI is more profitable than trading personally. This inevitably means that the money earned by AI is less than the total sum paid by all paying users. A counterargument to this is that AI can be seen as a tool that makes it easier for people to customize their AI, simpler than developing from scratch, and each person's AI version will be slightly different. However, I think this argument is somewhat weak and lacks convincing power.
Furthermore, a very successful AI algorithm, if used by many people, will see its effectiveness decline. If everyone in the market is using it, it won't be as effective. The market is a game of group psychology; you are essentially trading with everyone. If everyone is using the same strategy, the first person to use it usually wins, and the last person to use it may not make money, even if the strategy is the same. In this case, factors such as speed, performance, and other elements become crucial.
Overall, I believe that AI will be widely used by traders in their own unique ways.」
