OpenAI is crashing.

CN
BTCdayu
Follow
2 hours ago

OpenAI is Crashing

For decades, I have witnessed countless companies head towards collapse.

OpenAI is permeated with a sense of danger.

OpenAI issued a "red alert" in December.

Altman sent an internal memo to employees asking them to drop everything they were working on because Google's Gemini 3 is capturing the market.

Salesforce CEO Marc Benioff publicly announced he was abandoning ChatGPT after using it for two hours, opting instead for Gemini.

ChatGPT's traffic declined in November, marking the second month of month-over-month decline since 2025. Meanwhile, Gemini's monthly active users surged to 650 million.

This company, which was supposed to build general artificial intelligence, is unable to maintain the competitiveness of its chatbot.

But the real story is about money…

According to financial data released by Microsoft, OpenAI lost $12 billion in a single quarter. Deutsche Bank estimates that the company's cumulative negative cash flow will reach $143 billion before it becomes profitable.

Their analysts bluntly stated, "No startup in history has ever experienced such massive losses."

They burn $15 million a day just on Sora.

They spend $5 billion a year creating copyright-infringing memes.

Even Sora's chief engineer admitted, "The current economic situation is completely unsustainable."

This is a significant mathematical problem that no one wants to discuss:

To double the performance of these models requires five times the energy and money.

The low-hanging fruit has already been picked.

Now, any minor improvement requires exponentially more computing power, more data centers, and stronger electricity.

Reports indicate that OpenAI's large-scale training in 2025 failed to produce models better than previous versions.

The release of GPT-5 disappointed users. Many expressed that it was "disappointing" or even "terrible." Because users preferred the old model, OpenAI had to revert to the release of GPT-40 within 24 hours.

Altman had promised that GPT-5 would make GPT-4 feel "a bit embarrassed." However, users complained that it performed worse in basic math and geography.

Subsequently, they released GPT-5.1 and GPT-5.2.

Each time, the same complaints arose: too bureaucratic, too safe, too mechanical, too boring.

Talent loss has exacerbated the situation:

Chief Technology Officer Mira Murati. Left.

Chief Research Officer Bob McGrew. Left.

Chief Scientist Ilya Sutskever passed away.

President Greg Brockman. Stepped down.

Half of the AI safety team members left. Reports indicate that several executives accused of "psychological abuse" under Altman's leadership.

Now, Elon Musk has filed a lawsuit seeking $134 billion in damages.

A federal judge has just ruled that the case will go to a jury trial in April. There is substantial evidence that OpenAI's leadership had promised to maintain the nonprofit structure funded by Musk.

Based on these assurances, Musk provided $38 million in early funding. Now, he wants a share of the $500 billion valuation.

OpenAI calls it "harassment," but the judge disagrees.

I believe the following will happen next:

The AI hype cycle is nearing its peak.

The trend of diminishing returns can no longer be concealed.

Competitors are catching up.

The number of lawsuits is increasing.

OpenAI needs to generate $200 billion in annual revenue by 2030 to justify its projections.

A 15-fold increase in five years, while costs continue to soar.

Even Sam Altman admits that investors are "too excited" about AI.

His exact words were, "Someone is going to lose a lot of money."

If I were currently running a well-performing AI startup, I would look for an exit opportunity. Sell quickly while the heat is still on.

My view:

I would not touch stocks related to OpenAI at such a high valuation. The risks are simply too high.

If you have exposure to the "Big Seven" through investments in AI infrastructure projects, consider reducing your holdings. The gap between the promised revolution and the reality achieved has never been so wide.

Smart money is shifting towards industries where valuations truly reflect fundamentals.

The trading prices of small and mid-cap stocks are nearing a decade low compared to large tech companies, while earnings growth is only slightly lower.

The market can price risk, but it cannot price chaos.

OpenAI is like chaos wrapped in a $500 billion valuation.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到support@aicoin.com,本平台相关工作人员将会进行核查。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink