In a data-dense keynote, Alvin Graylin argues we are walking a razor's edge — not choosing between utopia and dystopia, but balanced on a ledge where the next five years will determine which side humanity falls. With AI models already scoring 140+ IQ and producing hours of autonomous work, the question is no longer whether jobs will be displaced but how fast the transition curve drops.

WHAT THIS PANEL COVERS

• AI models now score above 140 IQ (99.8th percentile of humans) and deliver 4-5 hours of autonomous work without human intervention — with full-day autonomy months away
• Descaling laws are replacing scaling laws: inference compute, distillation, and quantization are putting data-center-level intelligence into pocket devices
• A Google engineer used Claude to solve in one hour what her team of five spent an entire year building — and Anthropic built Claude Code Workbench in 1.5 weeks using AI writing AI
• AGI is deflationary for basic goods but inflationary for unique experiences and luxury — the money saved on cheaper food, energy, and medicine shifts to scarce human-centric value
• The biggest threat is not technology but mindset: social safety nets are affordable, humanoid robots are overhyped toys, and the real risk is governments and labs pursuing dominance over collaboration

SPEAKER

• Alvin Wang Graylin — Chairman, Virtual World Society; Global VP of Corporate Development, HTC; Author, 'Our Next Reality'; Digital Fellow, Stanford Digital Economy Lab

unDavos is a community-driven summit running during WEF week in Davos, democratizing the conversation around global challenges.

🌐 undavos.com

Tags: AGI, AI jobs, AI displacement, descaling laws, AI IQ, autonomous AI agents, humanoid robots, social safety net, AI bubble, AI commoditization, Alvin Graylin, Our Next Reality, Davos 2026, unDavos, REWIRE, future of work