A Drama About the “AI Doomsday”: Who Is Calling for a Stop, and Who Is Accelerating?
Hello everyone, I’m your financial journalist. Recently, something quite interesting happened in the tech world: several AI giants who usually have a lot of disagreements and even engage in legal battles have suddenly joined forces and declared in unison, “AI is too dangerous; we need to slow it down!”
Meanwhile, U.S. President Donald Trump loudly countered over the phone, “Don’t talk nonsense; this is a scam. We need to accelerate; otherwise, we’ll lose to China!”
At first glance, it seems like humans are discussing a matter of life and death. But if you peel away the layers of “ethics” and “safety,” you’ll realize that this is actually a carefully calculated power play. Today, we’ll break down this seemingly serious debate into five aspects to see who really has what interests at stake.
---
1. From Enemies to Allies? “Safety” Becomes the Most Valuable Marketing Tool
First, let’s look at an unusual phenomenon: Dario Amodi, the CEO of Anthropic; Sam Altman, the CEO of OpenAI; and Elon Musk, the founder of xAI.
What’s the relationship between them?
- Amodi and Altman: They are fierce rivals, with their companies constantly competing with each other.
- Musk and Altman: They are old enemies; Musk has even sued OpenAI, and they frequently insult each other online.
Logically, they should be working against each other. But this time, they surprisingly reached a consensus, both advocating for “slowing down AI” and emphasizing “safety.”
Why?
Because “safety” has become the most valuable marketing pitch in the AI industry.
- Anthropic (Amodi): This company was founded with the slogan of “putting safety first” and split from OpenAI. It’s about to go public (IPO) with a valuation of nearly a trillion dollars. To convince investors that their money is well spent, they need to emphasize, “We’re different; we’re not only intelligent, but we’re also safe. We’re responsible for humanity.” This “safety narrative” can directly boost its valuation.
- OpenAI (Altman): Although it also used to talk about safety, recently, to make money, it quietly removed the commitment to prioritize safety from its corporate documents and even lifted the ban on military use. Now it’s advocating for slowing down because it doesn’t want to go public right now. OpenAI lost $38.5 billion last year, while its competitor Anthropic has already turned a profit. If it goes public now, its valuation would likely be lower. By framing the delay as “a pause for human safety,” Altman saves face and buys time for a better valuation opportunity.
Conclusion: They’re calling for a slowdown not to save humanity, but to charge a higher price or avoid a setback.
---
2. The “Stalling Tactics” of the追赶ers: Musk’s Calculations
Elon Musk has also joined the “slow down” camp, but his motives are more subtle.
Musk has long said that AI is dangerous, yet he also founded xAI. His logic is practical: Since AI is inevitable, I’ll take control and make sure it develops in my way.
Currently, in the race for AI models, xAI is the chaser. OpenAI and Anthropic are moving faster and are more technologically advanced.
- If the industry races at full speed, the leaders will get even further ahead, and it will be difficult for xAI to catch up.
- If the industry slows down or stricter regulations and mutual testing mechanisms are introduced, the leaders’ advantages will be weakened, bringing everyone back to a more even starting point.
So, Musk’s proposal for “competitors to test each other’s models” is, on the surface, about industry self-regulation, but in reality, it’s about creating barriers for the leaders. He hopes to slow down his opponents by setting rules, giving himself more time to catch up. As one of the rule-makers, he can also gain more influence in the industry.
Conclusion: For Musk, “slowing down” is a tactical pause to reduce the gap with the leaders.
---
3. NVIDIA’s Anxiety: Why Jensen Huang Opposes Slowing Down?
In this debate, Jensen Huang, the CEO of NVIDIA, is a firm supporter of accelerating AI development.
NVIDIA is the “tool seller” in the AI industry: no matter which company discovers the “gold mine” (AI applications), they need to buy NVIDIA’s chips to process the data.
- The faster AI develops, the more data centers are built, and the more NVIDIA’s chips are sold, leading to better profits.
- If everyone says AI is too dangerous and needs to be regulated, data center construction will slow down, reducing NVIDIA’s orders and dragging down its stock price.
Therefore, Huang strongly opposes the “AI doomsday” theory, claiming that predictions of AI replacing humans or jobs have not come true. He needs to maintain the narrative that AI is experiencing exponential growth because his company’s performance depends on this narrative.
Conclusion: Huang opposes slowing down because it means fewer chip sales and less profit.
---
4. Trump’s Electoral Calculations: Labeling “AI Risks” as a “Scam”
President Trump’s role in this drama is crucial. He not only opposes slowing down but also directly called Huang, stating publicly, “Robots won’t take over the world; this is a scam.”
Trump’s actions are driven by two main reasons:
1. Sino-U.S. Competition: Trump has always emphasized that the U.S. leads China by only one year in AI, and this advantage could be lost at any moment. If he slows down due to concerns about “AI getting out of control,” it would mean handing over the advantage to China. For him, speed is a matter of national security.
2. Midterm Election Pressure: This is a more practical factor:
- Americans are quite dissatisfied with AI. Why? Building data centers consumes a lot of electricity, leading to higher energy costs; many also worry about AI taking away jobs.
- Polls show that 70% of Americans oppose building AI data centers locally, and support for AI is even lower than support for the Immigration and Customs Enforcement Agency (ICE).
- If Trump supports slowing down AI or stricter regulation, it would imply that there are indeed issues with AI, which would harm tech stocks and make voters think the government is overreaching.
- Therefore, he must label calls for regulation as “absurd conspiracy theories” and “scams” and promote data centers as “the future’s oil” to appease voters and maintain his support base.
Conclusion: Trump opposes slowing down to win the election and preserve the U.S.’s technological advantage over China.
---
5. The Truth: No Doomsday, Only Business
Let’s summarize the essence of this debate.
As the article says, “If something could truly end humanity within decades, debates wouldn’t solve it. If it were solvable through debate, it wouldn’t be called a doomsday in the first place.”
Current AI, although powerful, is still far from the sci-fi scenario of “destroying humanity.” Claims of AI taking over the internet within 6-12 months or causing billions of dollars in losses are more abstract threats with no concrete evidence to support them.
On the contrary, these companies are all shouting about “safety” while quietly removing safety measures to speed up their profit-making:
- OpenAI has removed its commitment to safety and started taking military orders.
- Anthropic has changed its strict rules to allow for more flexible reporting on risks.
- Everyone is rushing to release new models to gain market share.
In essence, this “AI doomsday” debate is a game of power distribution:
- Companies looking to go public (like Anthropic) need the “safety” narrative to boost their valuation.
- Loss-making companies (like OpenAI) need it to delay their IPOs.
- Chasers (like xAI) need the “slowing down” narrative to slow their opponents.
- Hardware sellers (like NVIDIA) need the “acceleration” narrative to sell their chips.
- Politicians (like Trump) need the “anti-slowdown” narrative to win elections and compete with China.
A lesson for the public:
Next time you hear news about AI destroying the world or needing regulation, don’t panic or be easily moved. First, see who is speaking and what their interests are.
- If they’re trying to sell you “safety solutions,” just listen.
- If they’re trying to slow down competitors, just observe.
- If they’re trying to make you unworried and continue spending on electricity and chips, be cautious.
The development of AI won’t stop, but its direction is determined by those with ulterior motives, not by philosophers.