← Retour au blog
tech 9 September 2026

I Resigned from Anthropic Today: A Look at the AI Race

Jacob Coxon leaves Anthropic, criticizing the reckless race towards superintelligence. Is the AI industry playing with fire?

Article inspired by the original source
I resigned from Anthropic today ↗ twitter.com

Introduction

Jacob Coxon, an influential researcher in the field of artificial intelligence, announced on X (formerly Twitter) that he is leaving his position at Anthropic. After three years of working on pretraining research at both OpenAI and Anthropic, Coxon is sounding the alarm about these companies' practices. According to him, they are rushing headlong into creating self-improving superintelligences, thereby gambling with humanity's future.

The Race to Superintelligence

The quest to develop a general artificial intelligence (AGI) has become the Holy Grail for many tech companies. OpenAI and Anthropic have positioned themselves as leaders in this field, investing heavily in the research and development of increasingly powerful AI models.

According to a recent PwC study, the global AI market could be worth $15.7 trillion by 2030. This economic prospect is driving companies to accelerate their efforts, often at the expense of ethical and safety considerations.

The Risks of Self-Improvement

One of the biggest concerns raised by Coxon is the risk associated with superintelligent systems capable of self-improvement. These systems could potentially surpass human intelligence, posing unprecedented ethical and security challenges.

Elon Musk, co-founder of OpenAI, warned back in 2015 that AI represented "the biggest existential threat" to humanity. Since then, debates have intensified on how to regulate and manage these powerful technologies.

Critiques of OpenAI and Anthropic

OpenAI, initially founded with the mission of ensuring that AGI benefits all of humanity, has been criticized for its shift towards a more commercial model. Similarly, Anthropic, although positioning itself as a company focused on AI safety, is not immune to criticism.

Researchers and experts, like Coxon, point out that the pressure to be the first to reach technological milestones could lead to dangerous compromises on safety.

A Call for Responsibility

In his message, Coxon calls for deeper reflection and collective responsibility in the development of AI. He advocates for stricter regulation and increased transparency in the research conducted.

Initiatives like the "Partnership on AI," which brings together tech companies, researchers, and NGOs, aim to establish ethical standards for AI. However, their effectiveness remains limited in the face of rapid technological advancements.

Conclusion

Jacob Coxon's resignation from Anthropic highlights the critical challenges and issues of the race to superintelligence. While the economic and technological benefits of AI are undeniable, the question of how to manage them responsibly remains open.

Let's discuss your project in 15 minutes.

intelligence artificielle superintelligence Anthropic OpenAI éthique IA
Deepthix newsletter · 100% AI · every Monday 8am

An AI agent reads tech for you.

Our AI agent scans ~200 sources per week and ships the best articles to your inbox Monday 8am. Free. One click to unsubscribe.

Visit the newsletter page →

Want to automate your operations?

Let's talk about your project in 15 minutes.

Book a call