Google Labs News

NO

YES

NO

The top Elo score on the EQ-Bench Creative Writing v3 leaderboard exceeds 2000 before 2027-01-01

At least three major AI labs (xAI, OpenAI, Anthropic, Google DeepMind) release public multi-LLM agentic reasoning models before 2027-01-01

xAI makes an official public announcement claiming achievement of AGI before 2027-01-01

Anthropic makes an official public announcement claiming achievement of AGI before 2027-01-01

OpenAI makes an official public announcement claiming achievement of AGI before 2027-01-01

Google DeepMind makes an official public announcement claiming achievement of AGI before 2027-01-01

YES

Before 2026-04-01

Before 2026-02-01

Before 2025-12-01

YES

US

China

UK

EU

UAE

East/South-East Asia excl. China

NO

30 - Medium-Hard

35 - Tough

40 - Hard

45 - Super Hard

50 - Brutal

No, but transformers are still key (the breakthoughs augment upon transformers)

No, a paradigm shift is needed (e.g. JEPA, etc.

Yes: transformers scale the best

NO