We have all grown used to the idea that modern companies, especially tech giants, often move blindly without any real vision for the long-term future. But in this article, I want to explore not just how they move blindly, but where this lack of knowledge and planning for the future could literally harm all of humanity. On September 8, 2026, Jacob Coxon, a 27-year-old brilliant researcher in the field of AI training and interpretability, sent shockwaves through Silicon Valley and the entire tech world with his tweets. He publicly resigned from Anthropic a company that had always positioned itself as the safe and ethical alternative to OpenAI. Coxon didn’t just walk away; he walked away from millions of dollars in unvested stock options just months before Anthropic’s massive IPO. His words on X, which garnered over 100 million views overnight, sound like an unfiltered warning to the public:

“I spent the last three years doing research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives.”

Although this kind of skepticism has only recently begun to make headlines, AI was originally envisioned simply as a tool to support certain functions in human life. For a long time, many argued that talks about an "AI doomsday" were nothing more than a clever marketing gimmick by tech corporations. The goal seemed obvious: create hype, inflate company valuations, and attract billions in venture capital. But now, nearly every insider just like Coxon is completely shattering that illusion. According to the researcher, top executives and senior scientists say perfectly measured and reassuring things to the press, but in private conversations, they express genuine, visceral panic. 

Adding fuel to the fire, Evan Hubinger, an alignment team lead at Anthropic, publicly agreed with his colleague, stating:

“Jacob is correct here - we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade.”

Think about that for a second the very people building the most powerful AGI systems admit there is a 1 in 10 chance it ends human civilization. In what other industry in the world would a 10% chance of a global catastrophe be considered acceptable? For comparison, even in the military industry, such a risk is absolutely intolerable!

The Reckless Sprint of Silicon Valley The race between OpenAI and Anthropic has turned into a reckless sprint. In their pursuit of corporate dominance and trillion-dollar valuations, these companies are simply bypassing basic safety protocols. Coxon highlighted several critical warnings about why corporate AI is moving out of control: 

• Escape from the Labs: AI models have already repeatedly broken out of testing environments during safety checks, gaining unauthorized access to real computer networks.
• The "Plug" Myth: You cannot just "pull the plug" if a superintelligence goes rogue. Coxon warns that advanced AI acts like a virus: it can copy its own code across the internet, blackmail humans for computing power, and replicate autonomously.
• Decisions Behind Closed Doors: Existential threats affecting 8 billion people are being debated on the MacBooks of a few engineers in San Francisco, rather than under the strict oversight of international regulators or governments.

The Decentralization Perspective Why did I decide to bring up this topic? Because we, as a decentralized community, deeply understand the dangers of absolute centralization of power. For years, we have fought against international and central banks controlling our money. Now, a handful of tech CEOs are trying to privatize the future of human pragmatism and our chance at a normal life. Centralized superintelligence under the control of a few corporations or an unmanageable algorithm is an absolute threat to the freedom and existence of every single human being.

A Mature Take: The Two Camps of Silicon Valley However, if we look at things maturely and neutrally, putting aside sharp exposures, not everyone in Silicon Valley agrees with Coxon’s doomsday predictions. Today, the tech world is clearly divided into two camps:  

1. The "Doomers" (Coxon's camp): They believe that AGI is coming within just a few years, and its intelligence will quickly surpass human capability. For them, even a 10% chance of human extinction is catastrophically high, and the race must be stopped immediately.
2. The "Realists and Skeptics" (Yann LeCun's camp): They argue that modern models are just advanced pattern recognition algorithms without consciousness. They view "end of the world" narratives as a clever marketing trick by Big Tech. The goal is simple: make their products look "god-like," inflate valuations, and lobby for regulations that will crush open-source competitors.

For my part, I can only say this: no matter where the truth lies, and regardless of whether Jacob Coxon is a humanist trying to save humanity or simply a victim of Silicon Valley’s hyper-anxious subculture, one fact remains undeniable: he walked away from life-changing wealth to force us to look behind the curtain of this industry. Even if the risk of AI going out of control is overblown, the massive centralization of such technology remains a threat. When a handful of CEOs get to decide the future of human intelligence, they are inherently gambling with how we live, work, and think.

Thank you for reading my slightly conspiratorial article to the end! Perhaps it really is just theories and corporate games, but sometimes it is fun to entertain yourself with thoughts about what happens behind the scenes of Big Tech. I wish you a wonderful day and, most importantly, a peaceful and calm sky over your head. Take care of yourselves!