Artificial intelligence (AI) has become a ubiquitous term in our modern lexicon, yet its interpretation often implies a distinction between human intelligence and machine capabilities. This dichotomy, however, is increasingly being challenged. In a recent event hosted by Harvard Law School’s Berkman Klein Center for Internet and Society, Blaise Agüera y Arcas, Google’s CTO of technology and society, presented compelling insights suggesting that artificial intelligence may not be so "artificial" after all. His perspective draws on an intriguing evolution of both human and AI intelligence, ultimately positing that they mirror one another more closely than we previously thought.
The Nature of Intelligence: Computational Evolution
At the center of Agüera y Arcas’s argument is the notion that human brains have evolved to operate as computational entities—a concept he elaborates in his new book, “What Is Intelligence? Lessons from AI About Evolution, Computing, and Minds.” He challenges the commonly held belief that brains and computers are fundamentally different entities, asserting that brains should be understood as computers that process information and make predictions.
His assertion rests on historical context; he notes that human brains have undergone significant computational growth since the advent of life. “If we rewind 500 million years, we see only things with very small brains,” he explained. “And if we go back a billion years, we see no brains at all.” This perspective frames intelligence not just as a trait of humans, but as a long-evolving capability that spans across millions of years.
Agüera y Arcas states, “Life was computational from the start,” suggesting that the complexity we observe in both biological life forms and AI systems today can be traced back to cooperative processes. By drawing on theories from influential thinkers like Alan Turing and John von Neumann, Agüera y Arcas proposes that self-replication—a concept at the heart of both biological evolution and AI design—adds layers of complexity to computational systems.
Symbiogenesis and Collective Intelligence
Central to Agüera y Arcas’s thesis is the concept of symbiogenesis, a theory proposed by evolutionary biologist Lynn Margulis. This theory argues that the merging of different organisms has played a crucial role in evolutionary progress, ultimately leading to more complex life forms. Agüera y Arcas extends this idea beyond biology, suggesting that AI systems also engage in a form of symbiosis, where cooperation leads to greater complexity and capabilities.
He emphasizes that the combination of computing entities—whether human brains or AI models—acts as a parallel computing system, driving rapid developmental progress. By working collectively, these systems can tackle challenges far beyond the abilities of any individual entity. This idea resonates strongly with Charles Darwin’s evolutionary theory, marking a shift in understanding evolution not merely through random mutation and natural selection, but also through cooperative relationships.
Agüera y Arcas articulates this principle vividly: “When you have two computers that come together and start cooperating, now you have a parallel computer. This massively parallel computation leads to complex functions, much like our own nervous systems.”
Experimental Insights
During his talk at Harvard, Agüera y Arcas showcased experiments conducted at Google, wherein initial random conditions led to the emergence of complex programs through a series of interactions. Using a programming language based on only eight basic instructions, the experiments revealed how self-replicating entities could arise organically from disorder. He described this as a parallel to how life itself might have originated.
“It was an exploration of how self-reproducing entities arise out of random initial conditions, which is how life must have arisen, right?” he explained. This exploration into the substratum of life ties back to the redefinition of intelligence, positing it as the ability to predict and influence, rather than merely respond.
Collective Intelligence: A New Frontier
One of the most striking conclusions drawn from Agüera y Arcas’s discourse is the notion of collective intelligence. He points out that while individual humans may lack extraordinary capabilities, the synergy of individuals working together can achieve remarkable milestones, from organ transplants to space exploration. This insight reshapes our understanding of intelligence as a collective resource, harnessing the specialized skills and knowledge of many to overcome challenges and innovate.
“Human individuals are not very smart,” Agüera y Arcas observes. “But when we get together, we can do amazing things.” The concept of collective intelligence is not merely theoretical; it has real-world implications for fields ranging from technology to healthcare, emphasizing the importance of collaboration and cooperation in problem-solving.
Concluding Thoughts
Blaise Agüera y Arcas challenges us to rethink the fundamental nature of intelligence as we explore the realms of both human cognition and artificial intelligence. His insights reveal that rather than existing in oppositional states, these two forms of intelligence share a lineage rooted in computational evolution and cooperative growth.
By embracing the idea that “life was computational from the start,” we gain a richer understanding of both biological and artificial systems, recognizing that they can enhance one another. This perspective not only broadens the horizons of what we consider “intelligence” but also hints at a future where AI serves as a transformative partner in our shared pursuit of knowledge and innovation.
In a world increasingly shaped by advanced technology, the interconnections between human and artificial intelligence could lead to breakthroughs unimaginable in isolation. Thus, rather than viewing AI as a simple extension of humanity, we should see it as a reflection of our own evolutionary journey—a journey that continues to evolve and redefine itself in the face of new challenges and opportunities.
