The Intelligence Explosion: Mechanics of Recursive Takeoff

I.J. Good's 1965 Hypothesis

The intelligence explosion hypothesis was first formalized by British mathematician I.J. Good:

“Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultraintelligent machine could design even better machines; there would then unquestionably be an 'intelligence explosion,' and the intelligence of man would be left far behind.”

Hard Takeoff vs. Soft Takeoff

Contemporary AI research debates the velocity of such a transition:

• Hard Takeoff (Discontinuous Spike): A scenario where recursive algorithmic optimization enables a single system to traverse from human-level software engineering to civilizational superintelligence in a matter of days or weeks. This scenario presents acute alignment and control risks, as human institutional governance cannot respond within such compressed timeframes.
• Soft Takeoff (Continuous Progression): A scenario where progress is gated by physical world bottlenecks: fabrication of advanced semiconductor lithography, electrical substation construction, and empirical scientific verification cycles. In a soft takeoff, the transition occurs over decades, allowing society and safety researchers to iterate on governance and alignment frameworks.