In a thought-provoking discourse on the future of artificial intelligence (AI), Carl Benedikt Frey, associate professor of AI & Work at the Oxford Internet Institute, challenges the prevailing narrative surrounding the concept of an “ultraintelligent machine.” In a recent article, Frey critiques the notion that the advent of such technology could lead to “humanity’s last invention.” This assertion, originally put forth by mathematician I.J. Good in the mid-1960s, posits that once a machine capable of outperforming human intelligence is created, it will subsequently engineer an even more advanced intelligence, resulting in a rapid acceleration of innovation.
Frey highlights that while the idea has captured the imagination of many, including figures like Google DeepMind’s Demis Hassabis, the trajectory of innovation is far more complex than a straightforward progression of intelligent machines. He argues that the process of discovery is akin to a chain, reliant on each individual link. Historical events, such as the 1986 Challenger disaster, demonstrate how even minor failures—in this case, a small rubber seal—can derail highly advanced systems. This analogy serves to emphasize that the path to meaningful inventions is often obstructed by unforeseen bottlenecks.
For Frey, the emergence of artificial general intelligence (AGI)—which aims to perform any cognitive task—might expedite initial phases of medical research, yet it would face significant hurdles in navigating the complexities of clinical trials, scaling production, and securing regulatory approvals. He cautions that as automation increases, the human role will not vanish but will rather shift to managing these persisting challenges, where human judgment and practical know-how remain essential.
Frey also points to a critical shortcoming of the last-invention thesis: it implies that humans could be rendered unnecessary in the oversight of AGI. He argues that intelligence is not simply a measurable quantity; instead, it manifests differently in machines compared to humans. A robust AGI might excel in speed and pattern recognition, yet it could falter in handling rare or unique scenarios. This disparity suggests that the combination of human insight and machine efficiency will continue to yield superior outcomes.
Drawing on the example of AlphaGo, which famously triumphed over human player Lee Sedol, Frey illustrates that perceived dominance can be misleading. Research conducted in 2023 revealed that by directing top engines into unconventional configurations, even a novice player could defeat advanced algorithms. This emphasizes that systematic weaknesses persist, where human intuition and adaptability can provide significant advantages.
Furthermore, Frey addresses the assumption that all pertinent knowledge can be easily codified for AI systems. He cites the transformative impact of the Ford Model T, not just for its design but for its production methods. Ford’s operational strategies were embedded in the routines and culture of his workforce, rendering them difficult to replicate through mere blueprints. This knowledge, dispersed and often unspoken, is essential for the functioning of complex systems.
Critics of Frey’s viewpoint might advocate for the implementation of pervasive surveillance technologies like sensors and cameras to codify this elusive knowledge. However, he warns that such an approach overlooks the realities of human behavior and privacy regulations, such as the European Union’s General Data Protection Regulation (GDPR). Any attempt to harvest human know-how at scale would face significant legal and ethical obstacles, limiting the feasibility of this strategy.
Ultimately, Frey contends that while AGI may revolutionize certain aspects of discovery—making expertise more accessible and speeding up experimentation—the process of invention will still rely on human skills that cannot be easily digitized. Practical know-how, established routines, and the ability to navigate complex systems are irreplaceable components of successful innovation.
As AI technology continues to advance, Frey posits that the value in the market will shift toward those who can deliver tangible outcomes. Rather than diminishing the role of humans, the rise of intelligent machines will underscore the importance of human contribution, positioning individuals as critical bottlenecks in the innovation landscape. The conversation surrounding the future of AGI is not just about what machines can achieve, but also about how humans will adapt and thrive alongside these emerging technologies.
See also
Anthropic Wins Preliminary Injunction Against Pentagon, Legal Risks Remain Uncertain
Germany”s National Team Prepares for World Cup Qualifiers with Disco Atmosphere
95% of AI Projects Fail in Companies According to MIT
AI in Food & Beverages Market to Surge from $11.08B to $263.80B by 2032
Satya Nadella Supports OpenAI’s $100B Revenue Goal, Highlights AI Funding Needs
