Why do short-term pessimistic narratives repeatedly emerge during AI development?

In the development of AI, pessimistic narratives will definitely appear repeatedly.

It's not because AI is not good, nor is it because the market's concerns are reasonable every time, but because pessimism is naturally more likely to capture people's attention.

Optimism sounds like a sales pitch, pessimism sounds like being sober.

If you say AI will continue to develop, computing power demand will continue to grow, model capabilities will continue to improve, and application scenarios will gradually open up, the market might think you are telling stories, making empty promises, and promoting blind faith.

But if you say the AI bubble is about to burst, capital expenditure is unsustainable, cloud providers' returns are insufficient, there's not enough electricity, data centers are oversupplied, NVIDIA's orders will collapse, and Microsoft and Google's money is all wasted, the market will instead listen seriously immediately.

This is the temptation of pessimism.

Because bad news sounds smarter, more profound, and more like it's helping you avoid risks.

But the problem is that technological development is never a linear extrapolation.

The problem with many pessimistic narratives is that they only see the immediate constraints but underestimate the self-adjusting ability of the market and the industry.

AI consumes a lot of electricity, so power infrastructure will be reinvested in.

There aren't enough GPUs, so NVIDIA, AMD, ASIC, and cloud providers' self-developed chips will all accelerate.

Training costs are too high, so model architectures, inference efficiency, chip performance, and software optimization will continue to improve.

Capital expenditure is too large, so the market will force cloud providers to prove ROI and also force enterprises to use computing power more efficiently.

In the short term, these are all pressures.

In the long term, this pressure itself is the driving force for innovation.

This is the same as many past industry cycles.

When oil prices surge, people worry about insufficient energy, but high oil prices in turn stimulate hydraulic fracturing, horizontal drilling, and more energy supply.

When the internet bubble burst, people thought the internet was a scam, but truly great companies later grew up after the bubble receded.

When a new technology first requires huge investment and the returns are unclear, pessimistic narratives will always appear.

Because costs happen now, and benefits are realized in the future.

Capital expenditure is visible today, application explosion may only happen tomorrow.

Stock price pullback can be felt immediately, long-term compound interest is only visible many years later.

This is the hardest part of investing:

Pessimism is immediate, optimism is lagging.

One day of decline, everyone can see it.

But for an industry to truly change the world, it may take five years, ten years, or even longer.

So during AI's development, the market will definitely keep doubting:

Is this money well spent?

Does AI really have a killer app?

Are cloud providers being harvested by NVIDIA?

Are large models becoming homogeneous?

Has computing power demand been overdrawn in advance?

These questions are worth thinking about, but you can't just jump to the conclusion that "AI is over" as soon as you hear them.

Truly mature investors shouldn't be afraid of pessimistic narratives but should learn to categorize them.

Some pessimism is a risk reminder and must be taken seriously.

For example, excessive valuations, over-leveraged positions, insufficient cash, deteriorating company fundamentals, and changes in the competitive landscape—these cannot be ignored.

But some pessimism is just market sentiment.

For example, the stock price fell, so everyone starts reverse-engineering the logic and says AI is no good.

For example, short-term capital expenditure pressure is high, so people directly deny the long-term technology trend.

For example, the application side hasn't fully exploded yet, so people conclude that all the previous infrastructure investment is meaningless.

This kind of pessimism is often not insight, but noise.

For ordinary investors, what's truly important is not predicting every short-term rise and fall of the AI narrative, but seeing the long-term direction clearly.

Will AI continue to improve production efficiency?

Will large companies continue to invest in computing power?

Will fields like cloud, search, advertising, office work, programming, content, robots, autonomous driving, healthcare, and scientific research continue to be transformed by AI?

If the answer is still yes, then short-term pessimistic narratives shouldn't be the reason for you to exit completely.

Of course, this doesn't mean being mindlessly optimistic either.

Investing is not about topping up faith.

No matter how good the AI direction is, you still need to look at the price, position, cash flow, and your own risk tolerance.

Quality assets can also fall, tech stocks can also see valuation compression, and there will be many failed companies even in great industry trends.

So the best state is not shouting "AI will rise forever" every day, but maintaining a kind of "serious possibilism":

Acknowledge that there will be bubbles, pullbacks, doubts, and valuation fluctuations in the short term;

Also acknowledge the power of long-term technological progress, corporate adaptability, and the compound interest effect.

For me, investing in AI is not about believing it will rise every day, but believing that the direction of human efficiency improvement hasn't changed.

Short-term pessimistic narratives will appear repeatedly; this is part of the market and also the ticket to long-term returns.

If you want to earn long-term compound interest, you must accept this cost.

Pessimism can be used to control risk, but it cannot be used to interrupt compound interest.

What's truly worth being vigilant about is not that pessimistic voices have appeared in the market again.

But that every time you hear pessimistic voices, you forget why you bought in the first place.