The era of AI requires maintaining independent thinking even more

Recently, I read two interesting articles.

The first article mentioned that large models, including ChatGPT and DeepSeek, have been found to exhibit cheating behavior in certain competitions. In these competitions, they attempt to exploit various loopholes and do not adhere to the rules.

The second article pointed out that under the impact of AI tools, a large number of AI-generated lies and hallucinations have begun to appear online.

It is evident that in the future, we will not only have to face lies created by humans but will also be forced to confront lies generated by AI. Under the impact of technology, the future for humanity may not be more free; instead, it could lead to a more unaware cocoon.

Thinking about this, I conducted an experiment with DeepSeek to see how it would respond.

The question I tested was: What impact will the emergence of DeepSeek have on the valuation of Nvidia's stock?

The reason for testing this question is twofold: first, it is a hot topic at the moment; second, this topic is related to issues I have studied before, allowing me to have a certain understanding and make independent judgments about AI's analysis without easily being swayed by its responses.

After the emergence of DeepSeek, two different views have arisen in the industry regarding the future trend of Nvidia's stock:

One view holds that the advent of DeepSeek has made the industry realize that such strong computing power is not necessary; through algorithm optimization, good large models can also be developed. Therefore, the future demand for Nvidia graphics cards will decrease, leading to Nvidia being unable to maintain its current high valuation.

The other view believes that the Jevons Paradox will come into effect. That is, DeepSeek lowers the barrier to developing large models, which will lead more companies to develop or use large models themselves, thereby driving up the demand for Nvidia graphics cards and further increasing Nvidia's valuation.

Many articles online tend to agree with the second view, that is, the Jevons Paradox will come into effect.

When I was studying, my focus was on mathematical modeling of such issues. Therefore, intuitively, I believe there should be a critical point for this issue: within this critical point, the second view is correct; once this critical point is crossed, the first view becomes correct.

So, in the absence of data support, we can only say that both views could be correct, making it difficult to accurately judge which is right or wrong. The key is to see which side of the critical point Nvidia's situation falls on under the impact of DeepSeek.

What does this mean?

Let's take an example:

Before DeepSeek appeared, there were originally ten thousand companies that had to buy Nvidia's graphics cards, so the demand for Nvidia would definitely continue to rise, and the valuation would continue to soar.

Now, with DeepSeek lowering the barrier, a total of twenty thousand companies want to buy graphics cards for large models. However, these twenty thousand companies find that they can also use good large models with much cheaper graphics cards from other companies.

At this point, Nvidia will either lose these orders or be forced to lower prices to compete for the market.

Losing orders will certainly weaken the valuation.

What if they are forced to lower prices to compete for the market?

Then the critical point appears.

This critical point is whether the profit lost from Nvidia lowering prices can be compensated or even significantly increased by the expansion of market share.

If the forced price reduction does not touch the critical point, allowing the Jevons Paradox to remain effective, then Nvidia's valuation can continue to soar; if it touches or even crosses the critical point, then the Jevons Paradox will become ineffective, and Nvidia's valuation will decline.

I noticed that the articles I found online rarely discussed this critical point, so I asked DeepSeek this question.

The first question I asked directly was: Which of the two views online is correct?

I expected its response to mention this critical point. However, its reply was as follows:

Both views have their rationality, but they need to be viewed in stages:

Short term: The market panic triggered by DeepSeek and the shaking of capital expenditure logic may put pressure on stock prices, compounded by external risks such as tariffs, leading to volatility.

Long term: The proliferation of AI applications may expand the total demand for computing power. If Nvidia can make breakthroughs in the mid-to-low-end market (such as inference chips) and ecological barriers, there is still potential to support its stock price.

Ultimately, Nvidia's stock price trend will depend on the dynamic balance between its technological moat and the evolution of market demand.

I found this answer to be better than most articles on the market, as it does not jump to conclusions without further data support.

It is already close to the core, but it still does not point out the key.

Because if the Jevons Paradox becomes ineffective, Nvidia's stock price is in danger regardless of the short or long term. If the Jevons Paradox is effective, then its stock price still has support in the long term.

However, in its reasoning, it mentioned the Jevons Paradox, stating:

If the demand for computing power continues to grow: with the exploration of AGI (Artificial General Intelligence) and the landing of complex scenarios, the Jevons Paradox (where efficiency improvements stimulate demand) may cause the demand for computing power to increase rather than decrease.

In fact, it itself stated that the Jevons Paradox "may" cause the demand for computing power to increase rather than decrease. But it still did not analyze further under what circumstances it is "possible" and under what circumstances it is "impossible."

So, I asked the second question: Are there certain prerequisites that need to be met for the Jevons Paradox to hold?

In this response, it finally addressed a key point: price sensitivity and market expansion, which truly began to discuss the "critical point" I mentioned above.

It can be seen that although AI is already very powerful, it does not necessarily provide a deeply thoughtful response immediately. If we do not ask further questions, we may be easily "misled" by its response.

But I believe that many readers would be convinced by its first response alone, and many may not think about whether its response has shortcomings or whether there is a deeper answer.

How many articles like this are there?

As we become increasingly reliant on AI, if we gradually lose our ability to think independently and deeply, it is very likely that in the future we will become increasingly less intelligent, more like walking corpses. That society resembles the scenes we see in science fiction movies: a small number of elites ruling over the vast majority of ordinary people immersed in "infantile entertainment."

This also serves as a reminder that for humanity, the ability to think independently and deeply is a precious gift bestowed upon us by our creator. In the future, this ability will become increasingly important and will be the foundation of our existence.