“Once the machine thinking method had started, it would not take long to outstrip our feeble powers. There would be no difficulty in making machines experiment with different branches of mathematics and finding out which of them were the most profitable. At some stage therefore we should have to expect the machines to take control, in the way that is mentioned in Samuel Butler’s ‘Erewhon’.”
This chilling prophecy sounds like a line from a modern science fiction blockbuster. However, it actually originates from the very dawn of the computing age. The words belong to Alan Turing, the brilliant mathematician who cracked the Enigma code. He recognized the potential for artificial superintelligence long before personal computers even existed. Specifically, he foresaw a future where digital minds would inevitably surpass their human creators.
Most people know Turing for his wartime heroics or his tragic end. Yet, his thoughts on the future of AI remain some of his most profound contributions. He did not view computers merely as calculators. Instead, he saw them as embryonic minds capable of growth. This specific quote serves as a stark warning from history. It suggests that once we ignite the spark of machine thinking, we may lose the ability to control the fire.
Tracing the Origins of the Prophecy
Finding the exact source of this quote requires a journey through computing history. You cannot find this statement in Turing’s most famous academic papers. Rather, it remained hidden in his unpublished notes for years. The world finally saw these words in 1959, five years after Turing’s death. His mother, Sara Turing, compiled his papers into a biography titled Alan M. Turing.
She included a chapter on computing machinery that contained this specific passage. She labeled the section “Intelligent Machinery, A Heretical Theory.” This title suggests that Turing knew his ideas were controversial. He understood that suggesting machines could “outstrip” humans would unsettle his contemporaries. Consequently, these thoughts remained largely theoretical during his lifetime.
Later scholars helped pinpoint the precise moment Turing formulated these thoughts. Source For instance, author David Leavitt conducted extensive research for his 2006 biography, The Man Who Knew Too Much. Leavitt traced the text back to a lecture Turing delivered in Manchester. . Thus, we know Turing spoke these words in 1951.
The Context of 1951
To appreciate the foresight involved, we must look at the state of technology in 1951. Computers were massive, room-sized behemoths. They used vacuum tubes and magnetic drums. They possessed a fraction of the processing power found in a modern greeting card. Yet, Turing looked past these physical limitations.
He focused on the theoretical potential of the “machine thinking method.” He realized that physical hardware would improve. Therefore, the real constraint was not the machinery, but the software—the “thinking method.” He correctly predicted that once machines could learn, they would accelerate past us. This concept of recursive self-improvement is now a core concern in AI safety research.
Furthermore, Turing’s timeline is remarkable. He made this prediction when the concept of “software” barely existed. He envisioned machines experimenting with mathematics on their own. Today, we see this reality in systems that teach themselves to play chess or fold proteins. Turing saw the destination clearly, even if the road there was still unpaved.
Analyzing the “Feeble Powers” Argument
Why did Turing describe human powers as “feeble”? He was not insulting human intelligence. Instead, he was making a cold, logical comparison between biological and mechanical substrates. Humans face biological limitations that machines do not suffer. For example, we eventually die and take our knowledge with us.
Machines, conversely, are potentially immortal. If a machine learns something, it can instantly copy that knowledge to other machines. They do not need to spend decades in school. Consequently, their rate of evolution can far exceed biological evolution. Turing understood that this speed differential is the critical factor.
Additionally, he noted that machines could work tirelessly. They do not need sleep or breaks. They can experiment with different approaches to a problem 24 hours a day. Over time, this relentless optimization leads to capabilities we cannot match. Therefore, the “feeble” nature of our powers refers to our biological constraints, not our lack of creativity.
The Connection to Samuel Butler’s Erewhon
Turing explicitly references Samuel Butler’s novel Erewhon in his warning. This literary connection provides deep insight into his thinking. Butler published Erewhon in 1872. In the book, Butler describes a society that destroyed its machines out of fear. The citizens worried that machines would eventually evolve consciousness and enslave humanity.
By citing this novel, Turing bridged the gap between science fiction and scientific fact. He validated the fears Butler expressed almost a century earlier. However, Turing approached the problem as a mathematician, not a novelist. He saw the mathematical probability of machines taking control.
He argued that if machines optimize for “profitability” or success, they might view humans as obstacles. This aligns with modern theories about instrumental convergence. Basically, an AI might take control not because it hates us, but because we are in the way. Turing realized that “taking control” was a logical outcome of superior optimization capabilities.
Modern Resonances in AI Safety
Today, Turing’s 1951 warning resonates louder than ever. We now live in the era he predicted. Machines are indeed experimenting with mathematics and solving problems humans cannot crack. Consequently, leading AI researchers frequently revisit Turing’s words. They serve as a foundational text for the field of AI safety.
Professor Stuart Russell, a renowned computer scientist, highlighted this quote recently. Source He featured it prominently in his 2021 BBC Reith Lectures. Russell used Turing’s prediction to frame the current crisis in AI alignment. He argued that we are rushing toward the very future Turing warned us about. .
Russell points out that we still lack a solution to the control problem. We are building systems that are increasingly powerful. Yet, we have not figured out how to ensure they remain subservient to our “feeble powers.” The fact that we are still debating a warning from 1951 shows how difficult this problem is. Turing identified the trap, but we have yet to find the escape route.
The Inevitability of the “Machine Thinking Method”
Turing’s phrase “Once the machine thinking method had started” implies an irreversible process. He viewed the development of AI as a threshold event. Once we cross it, we cannot go back. The genie does not return to the bottle. This irreversibility makes the warning particularly urgent.
We see this momentum in the current tech industry. Companies race to build faster, smarter models. They compete for market dominance. As a result, the speed of development accelerates. We are fulfilling Turing’s prediction of machines finding the “most profitable” paths. The “machine thinking method” has definitely started.
Moreover, the integration of AI into global infrastructure increases the stakes. We trust algorithms with our stock markets, power grids, and defense systems. We are effectively handing over the keys, just as Turing predicted. The question remains whether we can maintain control as the systems grow more complex.
Recommended Reading & Resources
For further exploration of Alan Turing and related topics, here are some excellent resources:
- Alan Turing: The Enigma: The Book That Inspired the Film The Imitation Game – Updated Edition
- Alan Turing Decoded: The Man They Called Prof
- Alan Turing: The Mathematician Who Decoded the World
- Reflections of Alan Turing: A Relative Story
- Alan Turing: A Life From Beginning to End (World War 2 Biographies)
- The very short and very simple mathematical story of Alan Turing: The Boy who Taught Machines to Think
- Alan M. Turing: Centenary Edition
- Alan Turing: A Life from Beginning to End (Large Print Biography Books)
- Alan Turing (Little People, BIG DREAMS)
- Alan Turing: The Enigma Man
- Alan Turing (A Life Story)
- Extraordinary Life of Alan Turing
As an Amazon Associate, we earn from qualifying purchases.
Conclusion
Alan Turing’s 1951 warning stands as a monument to human foresight. He looked at primitive vacuum tubes and saw the potential for a new dominant species. His quote about machines outstripping our “feeble powers” is not just a historical curiosity. It is a defining challenge for our generation.
We must take his reference to Erewhon seriously. The transition of power from biological to mechanical intelligence is not guaranteed, but it is possible. Turing gave us the map and the warning signs. Now, it is up to us to navigate the territory safely. We must ensure that while machines may outstrip our processing speed, they never outstrip our values.