Humanity's Last Invention

In 1965, statistician I. J. Good sketched a thought: if we ever built an ultraintelligent machine, it could design even better machines, triggering an “intelligence explosion.” In that scenario, “the first ultra-intelligent machine is the last invention that man need ever make.” It’s a line that is both a prophecy and warning.

Good wasn’t a science fiction writer, he was a Bayesian statistician who worked with Alan Turing at Bletchley Park during World War II helping crack the Nazi Enigma code. When he wrote about intelligence explosion, he was extrapolating from mathematics, not imagination. His formulation was precise: a machine slightly smarter than humans could design a machine slightly smarter than itself, which could design something smarter still, accelerating until human intelligence became a distant baseline.

Sixty years later, here we are, with the November 2025 release of Gemini 3.0 achieving a 130 or “genius” level on the standardized IQ test (“smarter” than 98% of humans). And today’s AI doesn’t just execute tasks; it is already helping build the next generation of AI. All of the large “frontier” model companies are using their current best models (better than the ones in public use) to build the next generation. Even beyond the models themselves, OpenAI has already demonstrated how AI systems can optimize chip designs for running AI. We’re right at the beginning of the intelligence explosion, with tools making better tools, capabilities compounding quarterly rather than generationally.

Good’s quote matters here not because we stand on a cliff, but because it names a direction of travel: tools that increasingly help build the next tools and bring about first “artificial general intelligence” (AGI) and then “artificial super intelligence” (ASI). The question for working adults isn’t “Will a machine replace me tomorrow?” It’s “How do I become the kind of human who knows how to shape and work with machines to achieve even better outcomes?”

The word that changes everything is “hybrid.” Good’s quote seems to imply a clean handoff in which humans invent the ultraintelligent machine, then step aside. But what we’re actually building is messier and more interesting: systems where human judgment and machine capability interweave so tightly you can’t separate them.

Consider what’s happening right now in every industry:

• Engineers don’t just use AI to write code; they use AI to architect systems that no single human could hold in working memory
• Designers don’t just generate images; they develop aesthetic languages that emerge from the conversation between human taste and machine possibility
• Strategists don’t just analyze data; they explore solution spaces that only become visible through human intuition plus data sets so large that only a machine could navigate them

This isn’t human versus machine. It’s human-machine systems growing more sophisticated together. The intelligence explosion Good imagined assumed machines working alone. What we’re getting is more interesting: hybrid intelligence that neither human nor machine could achieve independently.

Here’s what Good got right that most people miss: the phrase “last invention that man need ever make” doesn’t mean humans stop inventing. It means the nature of invention fundamentally changes.

When tools can improve themselves, human creativity shifts from direct creation to something more like cultivation or conducting. You’re not building the thing; you’re directing and shaping the thing that builds things. You’re not writing the code; you’re designing the intention that generates code. You’re not creating the strategy; you’re setting the values that guide strategic exploration.

This shift is already visible in how the best practitioners work with AI:

• They don’t write prompts; they design prompt systems that evolve
• They don’t generate outputs; they create evaluation frameworks that select and refine outputs
• They don’t use tools; they build tool-chains that compound capabilities

The “last invention” is a phase transition in what invention means.