Éric Schmidt: 'In Six Years, AI Will Be Smarter Than All Humans Combined'

Former Google CEO Éric Schmidt issues a shocking warning: artificial intelligence could surpass the sum of human intelligence within six years. According to him, current AIs are evolving towards a form of 'self-improvement' that enables them to enhance themselves faster than humans can, posing major challenges to society.

An Acceleration Beyond Human Control

An AI Capable of Self-Enhancement

In a recent interview, Schmidt highlighted that computers now have the ability to reorganize, learn, and plan without human intervention. He mentions an intelligence capable of escaping human control, both in speed and complexity.

It's important to grasp the significance of this statement. Schmidt claims these developments are not distant; they are part of an already initiated trajectory. Today's AIs could be the architects of their own future versions—an exponential improvement cycle.

A Break Before the Intelligence Explosion

For Schmidt, it's a matter of six years. At corporations like Schmidt Futures or MIT, he warns that AI will no longer be a tool: it could become a self-evolving and autonomous entity. According to him, this form of "super-human" intelligence could surpass the sum of all human intelligences combined, ushering in a new era.

Ethical, Geopolitical, and Governance Challenges

A Challenge for Regulation and Ethics

The ultra-rapid development of artificial intelligence questions the very principles of regulation. AI could become too complex for traditional legal frameworks to fully contain: how to regulate an intelligence that modifies itself?
The risks particularly concern algorithmic trust, autonomous decisions in sensitive areas (health, armament, justice), and process traceability.

The Global Strategic Issue

The advancement of AI is also a geopolitical race. Schmidt warns against an asymmetric technological battle: who will control these super-intelligences?
As the United States and China focus their efforts on these technologies, Schmidt calls for international cooperation: without concerted global governance, the risks of abuses—espionage, mass manipulation, economic disruption—could become unmanageable.

Éric Schmidt predicts a silent but imminent revolution: an AI capable of surpassing human collective intelligence within six years. This vision demands urgent reflection: how to define fair and effective rules to frame a technology that learns and improves on its own?

To frame this evolution, it is crucial to:

• establish obligations for transparency, algorithmic traceability, and third-party reviews;
• implement regular independent audits to assess the self-evolving capabilities of AIs;
• formalize kill switch clauses in case of unforeseen or dangerous behaviors.

The question is no longer whether AI will surpass human intelligence, but how we collectively prepare. Establishing robust international standards to frame machine autonomy could well become the civilizational challenge of the 21st century.