Demis Hassabis, CEO of Google DeepMind, used Google I/O's keynote to declare the present moment a watershed for artificial intelligence. He positioned current advances in AI research and Google's products as stepping stones toward artificial general intelligence, or AGI, the theoretical point where AI systems match or exceed human cognitive abilities across all domains.
Hassabis framed this period as the "foothills of the singularity," invoking the concept of technological singularity, a speculative threshold where AI improvements accelerate beyond human comprehension or control. The phrase signals that, in his view, transformative capabilities already exist today, but the most dramatic changes remain ahead.
The statement carries weight coming from someone who led DeepMind through breakthroughs in game-playing AI and protein structure prediction. Google I/O typically showcases near-term product applications, but Hassabis pivoted toward longer philosophical terrain, linking immediate research wins to existential technological shifts.
The rhetoric matters because it shapes how the industry, regulators, and the public perceive AI development. Framing current progress as approaching AGI tightens the narrative around Google's competitive position and justifies heavy R&D investment in foundational models. It also echoes a common Silicon Valley pattern: grounding near-term products in far-reaching, world-changing ambitions.
However, no consensus exists on whether AGI is genuinely close or decades away. Researchers debate whether current language models and AI systems possess genuine reasoning or simply pattern-match at scale. Hassabis's language sidesteps these technical questions, instead emphasizing moral responsibility: unlocking AGI's potential "for the benefit of the entire world."
The timing aligns with intensifying competition between Google, OpenAI, Anthropic, and other labs racing to build more capable systems. By characterizing this moment as historically