Recently, I noticed an interesting phenomenon: the stock price trends of U.S. artificial intelligence listed companies are essentially repeating the same story—since ChatGPT became popular, the entire AI concept stock market hasn't settled down.

Honestly, the investment opportunities in this wave of AI are indeed huge. According to the latest data, global corporate spending on AI solutions has surpassed $300 billion, and this number is growing at nearly 30% annually. By 2028, the total AI industry expenditure is expected to double, which means there is still considerable room for future growth.

I’ve observed a key phenomenon: the companies that are truly making money are not those loudly promoting AI concepts, but those providing underlying infrastructure. Just like during the gold rush, selling shovels earned more than digging for gold. This also explains why chip manufacturers like NVIDIA have become the most watched targets by global investors.

NVIDIA’s situation is particularly noteworthy. The company's GPU chips have become the industry standard for AI training and inference, with a market value surpassing $4 trillion, and its stock price has increased more than tenfold compared to before ChatGPT appeared. Their moat is very deep—if AI development can survive without anyone, then it can’t survive without NVIDIA’s chips. Most importantly, current production capacity still lags behind demand, which means future growth potential is even greater than it appears now.

Besides NVIDIA, Broadcom is also a big winner in this AI wave. Through years of acquisitions, this company has nearly monopolized the network communication chip market. The development of AI relies on high-speed networks; whether it’s the chips themselves or data transmission, their products are essential. In fiscal year 2024, AI-related product revenue accounted for 25%, with a very strong growth trend. In less than two years, Broadcom’s stock price has increased by 3.5 times.

Although AMD still lags behind NVIDIA in the GPU field, their MI300 chips are already competitive in many performance tests, and the key is that their price is only half of the H100. This gives many companies a second option. As AI application scenarios become more diverse, the demand for alternative solutions is indeed increasing. AMD’s data center business growth has reached 27%, a trend worth watching continuously.

Investing in U.S. AI listed companies is actually quite flexible. Buying individual stocks is the most direct approach, but if you want to diversify risk, you can also consider AI-themed ETFs. I’ve seen some investment institutions adopt a strategy of dollar-cost averaging combined with portfolio adjustments, which allows participation in AI growth dividends without being too affected by the volatility of a single company.

However, a reminder here: although AI concept stocks have promising prospects, they also carry significant risks. First, this field changes too quickly; the benefits seen this year could disappear next year. Second, many AI companies already have high valuations, and some even lack clear profit models. Third, policy and regulatory changes could suddenly alter the game.

In the long term, the beneficiaries of this AI wave are likely to be concentrated in two types of companies: one is providers of chips and servers and other infrastructure, and the other is enterprises that can truly implement AI technology into practical applications. In the short term, there may still be volatility, but the overall trend should be upward. The key is patience—don’t be scared by short-term fluctuations, and avoid blindly chasing highs. Regularly review your holdings and adjust your strategy according to market changes so you can achieve relatively stable returns from this AI dividend wave.