Citrini: Investing in AI infrastructure makes the future trend of the insurance market unpredictable

CitriniResearch released its latest view that the current results of everything related to AI are super uncertain, and the impact on the future may be extremely different. A little assumption that is casually changed may lead us into a completely different world tomorrow. This uncertainty is not only in hardware such as computing power, chips, and data centers, but also in the question of "whether there will be a serious shortage of memory in the next ten years." How consumers will use AI and what the entire AI market will look like ten years later are also unclear. For example, Anthropic may go from the underlying infrastructure to the top-level application software and then dominate the market; it is also possible that major AI laboratories have been "cabbage", and the ones who really make money are cloud vendors, specific application developers, or those who have not yet emerged. New innovators. When CitriniResearch was optimistic about AI in early 2023, it decided to only invest in "pickaxes and shovels"(referring to infrastructure) because we couldn't guess what AI would look like ten years later. But as long as AI is to develop, it will definitely need more hardware and computing power, which is relatively safe. Three years have passed, and I thought that the future direction would become clearer and clearer, but the result became clearer and clearer. But this is not entirely a bad thing-it means that there are still many opportunities, and everyone in the wave needs to be flexible and able to keep up with new changes. Tools related to agent AI have performed well, but no one knows what form AI will actually be implemented in five years.