Deepmind researchers propose "Artificial Symbiotic Intelligence" as an alternative to the singularity
decoder ·
DeepMind researchers Benjamin Bratton, Blaise Agüera y Arcas, and James Manyika propose "Artificial Symbiotic Intelligence" as an alternative to the traditional concept of a technological singularity. They argue that artificial general intelligence will emerge from social systems where humans and AI agents collaborate, rather than from isolated superintelligent machines. This perspective is supported by observations that reasoning models naturally develop multi-perspective behaviors during training without explicit programming. The authors suggest that as AI agent instances grow while human populations shrink, society may reach a cognitive tipping point similar to the Industrial Revolution. In this new paradigm, AI agents are viewed as temporary, decomposable assemblages of tools and roles rather than entities with fixed identities or continuous selves. Consequently, future interfaces may shift from one-on-one chats to visual network diagrams that allow users to coordinate swarms of agents. The researchers emphasize that governance and institutional frameworks are crucial for managing these complex interactions, as markets alone cannot handle social nuances like virtue or guilt. Ultimately, alignment is reframed as an ongoing negotiation shaped by institutions rather than a static imposition of values on individual models.
decoder