What happened
A 57-page technical report authored by 14 Google DeepMind researchers — including co-founder Shane Legg and Marcus Hutter (creator of the AIXI universal intelligence framework) — maps the transition from human-level AGI to artificial general superintelligence (ASI), submitted to arXiv on June 10, 2026 (preprint, not peer-reviewed). The report characterises ASI as "a system that is more intelligent and cognitively capable than large organisations of humans" and identifies four non-mutually-exclusive pathways to get there: scaling AGI, AI paradigm shifts, recursive self-improvement, and ASI emerging from large-scale multi-agent collectives. It then systematically analyses potential frictions and bottlenecks along each pathway and concludes that "it cannot be ruled out that AI progress might continue to accelerate over the next years," warning that the popular image of a single transformative AGI step-change "could be inaccurate" — more apt is "a series of transformative societal changes." The paper sets out a concrete interdisciplinary research agenda of open questions required to bound these uncertainties.
Why it matters
This is the most authoritative attempt by a frontier AI lab's own scientific leadership to formally bound what comes after AGI — a question boards and governments are increasingly forced to plan around. Executives relying on the "AGI as a single event" model for scenario planning should recalibrate to the multi-shock, parallel-pathway picture this report presents.
Action needed
Assign a senior strategist to read Section 5 (Technological Pathways and Potential Bottlenecks) and map the four ASI pathways to your organisation's AI risk taxonomy and 3-5 year scenario planning assumptions.