Ordinary people buying ETFs should not immediately follow the trend AI has ignited many hard tech ETFs, from semiconductors, storage, lithography, to power grids, nuclear power, and robotics, all with their own logic But I think the order in which ordinary people buy ETFs is very important The first layer is still wide base For example, VOO, VTI, VT, or if you want to lean towards technology, look at QQQ/QQQM These are more suitable for long-term bottom positions On the second level, let's take a look at the theme of mature technology If you are optimistic about AI, semiconductors will definitely not be able to bypass it ETFs such as SMH and SOXX are more mature than many new concept ETFs and are more suitable as core observation objects in the AI hardware direction The third layer is the more segmented AI related topics for example DRAM: biased towards storage and HBM, logic is that AI requires stronger memory and bandwidth EUV: leans towards advanced processes, lithography, and semiconductor equipment, but should not be simply understood as "only buying ASML" GRID: Power grid upgrade. AI data center expansion, power and grid will be long-term constraints NLR/URA: biased towards nuclear power and uranium industry chain, logic is to stabilize the rising demand for electricity BOTZ/HUMN: robot direction, BOTZ is more mature, HUMN is earlier NASA/ARKX: More focused on space economy and commercial aerospace, with intersections with AI, but not including core AI infrastructure, with greater fluctuations FOTO: Polarized Communication and Photonics Technology, belonging to a more segmented branch of AI infrastructure I personally think that ordinary people don't need to buy all ETFs at once. Buying too much may seem scattered, but in reality, they may all be the same technological growth risk If you really want to buy, Kuanji is still the main position. Semiconductors can be equipped with a little bit, while power grids, nuclear power, storage, and lithography depend on personal preferences and can be added in small quantities. Space, humanoid robots, photons, and the like are mostly observation chambers. Don't rush into heavy storage just because you're excited ETF is not a tool for making profits, nor is it a way to buy and not worry about it Theme ETFs should particularly consider their fees, size, trading volume, establishment time, and top ten holdings The right direction doesn't mean buying something will always be right A good track does not mean that every ETF is worth investing heavily in Ordinary people buying ETFs should not expect to seize every opportunity. Comfortable storage space and the ability to hold onto it for a long time are actually more important