Hugo de Garis – AI, Humanity, and the Longer Term
About | Start Time | How to Join
Hugo de Garis will give a presentation at Future Day 2025.
About
Abstract:
Hugo de Garis believes that too many commentators on AI are avoiding the fundamental issue: before long, machines will become vastly more intelligent than humans—potentially trillions of trillions of times more, or even beyond that. Humanity will soon face a critical decision: either accept that humans will become the second most intelligent species or impose a global ban on the creation of artilects (artificial intellects).
Most AI researchers lack the mathematical and physics background necessary to fully grasp the enormous potential for godlike intelligence that future artilects could possess. For instance, consider a grain of sand enhanced through nanotechnology to process one bit per atom. If each atom switches states in femtoseconds, the total processing capacity of that grain of sand would be a quadrillion times greater than the estimated processing power of the human brain. And this assumes purely classical computation—whereas a quantum computer, capable of processing 2^N operations simultaneously, could achieve truly godlike computing capacities.
Sooner or later, humanity will have to answer the ultimate question: Do we build gods, or do we create our own potential exterminators?
Given that the stakes involve the very survival of the human species, the debate over this issue will become increasingly intense as machines approach human-level intelligence. The resulting ideological conflict could escalate into the most devastating war in human history—the “Artilect War”—fought with 21st-century weaponry, potentially leading to the deaths of billions: “giga-death.”
Bio
Hugo de Garis (born 1947) is an Australian retired researcher in the sub-field of artificial intelligence (AI) known as evolvable hardware. He became known in the 1990s for his research on the use of genetic algorithms to evolve artificial neural networks using three-dimensional cellular automata inside field programmable gate arrays. He claimed that this approach would enable the creation of what he terms “artificial brains” which would quickly surpass human levels of intelligence.
See Also
