2026 投资指南【结尾附详细股票清单】

Many people now think that AI has already risen too high, but I don't think so. The impact of AI on humanity is just beginning to show, much like the products of the steam revolution and the industrial revolution—steam engines widely used in production equipment, trains, cars, etc.—which brought earth-shaking changes to people's lives.

Or take the information revolution, which you could also call the internet revolution, with all kinds of computers, smartphones, etc., leading to an explosion of information and similarly earth-shaking changes in people's lives.

Now, looking at the present, we call it the AI revolution. How much has your life been affected? Are robots running all over the streets? I forget whether it was Musk or Jensen Huang who said in an interview that AI won’t leave you with nothing to do; instead, it will keep you so busy that you can’t keep up. In the past, it took a long time to handle one task, and you could only handle one thing a day. But now, with AI’s help, you can handle 100 tasks a day. Your workload actually increases, which shows that AI brings higher efficiency and stronger productivity. Suppose the world used to solve 60 billion tasks a year at its peak; with AI, it can solve 600 trillion tasks. Think about the value created by solving all those tasks.

Right now, AI is still involved in too few things globally, and many people haven’t been affected by it. When robots are everywhere and AI is omnipresent, then you can say there might be a bubble and things are a bit overvalued.

If we divide each revolution into early, middle, and late stages, the current AI revolution is at most in its early stage.

If you can’t see a single robot on the streets, it doesn’t make sense to say we’re already at the end. Musk is still building chip factories because he believes TSMC, no matter how advanced, won’t be able to meet his future demand—he has to produce his own.

The development of AI will iterate very quickly. The more powerful AI becomes, the more chips it will need because higher efficiency means more tasks to handle. For example, in a company, a highly capable person can do in one hour what others take a whole day to do. Naturally, the boss will assign more tasks to them, so their workload doesn’t decrease—it actually increases. The same logic applies here.

Therefore, in the early stages of AI, you shouldn’t bet on doomsday scenarios. Avoid being thrown off by market fluctuations.

A balanced approach is key: steady investments, long calls, and diversified development.

Steady investments provide stable returns, long calls accelerate gains, and going all-in at the right time maximizes profits. Combining these three strategies ensures both offense and defense, balancing returns and controlling risks.

Diversify across multiple sectors to avoid excessive risk.

AI + defense + commercial aerospace + pharmaceuticals (balanced allocation).

If you divide your capital into 100 parts, here’s a reasonable distribution. The list below consists entirely of established players with high growth potential and unlimited upside. Adjust as needed!

$Western Digital(WDC.US) 10 parts (steady 5, long call 3, all-in 2)

Investment rationale: AI will lead to endless data stacking. The demand for storage capacity far exceeds that for computing power and grows exponentially. If computing power is like a machine that produces cans, storage capacity is more like the cans themselves. The higher the machine’s efficiency, the more cans it produces, and the more cans you need. Here’s an example: One machine can produce 100 cans a day. Add one machine every day, and after a year, you’re producing 36,500 cans a day. Once you buy the machines, as long as they don’t break, they’ll keep producing. So even if you stop buying new machines, the existing ones will still need a huge number of cans. This means that even if computing chips stop selling, storage will still be needed because all those machines are still running nonstop!

$Tesla(TSLA.US) 10 parts (steady 5, long call 3, all-in 2)

$Lockheed Martin(LMT.US) 10 parts (steady 5, long call 5)

Investment rationale: America’s hegemonic position is becoming increasingly unstable, with internal conflicts surging. The enormous fiscal burden is making it hard for this war machine to keep up. The rise of AI will inevitably lead to the iteration of weapons, and more powerful weapons will be produced quickly with AI’s help. The root cause of every world war is that after capitalism reaches the monopoly stage, conflicts over markets, raw materials, etc., are ultimately resolved through war. The internal pressure in the U.S. is already obvious. Legitimate business alone can’t solve such a huge fiscal burden—plunder is the only way out. Plunder requires stronger weapons. Although the U.S. is currently the top military power, it has also made countless enemies. If it really wants to take on the whole world, it will face military pressure, but it has no choice. Therefore, further military expansion is urgent.

$Novo Nordisk AS(NVO.US) 10 parts (steady 5, long call 2, all-in 3)

$Intel(INTC.US) 10 parts (steady 3, long call 5, all-in 2)

Investment rationale: Even Musk is building his own chip factory. Although Intel, as an old-school company, isn’t as good as TSMC in chip production, it’s still the home team. No matter how good TSMC is, it’s not a domestic brand. The U.S. would much rather have its own brand take the lead than rely on foreign companies. So, under these circumstances, the U.S. will provide all kinds of support to Intel, including funding and even technology. The U.S. is already racially discriminatory toward its own people—if it weren’t for the lack of alternatives, it wouldn’t let foreign companies make all the money! It’s inevitable that Intel will take some of TSMC’s foundry orders in the future. Plus, while many stocks have soared in the AI wave, Intel has barely moved, rising just a tiny bit 🤏. It’s a value play with unlimited long-term potential!

$XTALPI(02228.HK) 10 parts (steady 5, turbo 5)

$Alphabet - C(GOOG.US) 10 parts (steady 4, long call 5, all-in 1)

$Taiwan Semiconductor(TSM.US) 10 parts (steady 2, long call 6, all-in 2)

$Rocket Lab(RKLB.US) 10 parts (steady 2, long call 8)

$Micron Tech(MU.US) 10 parts (steady 5, long call 3, all-in 2)