Purpose
Document the transition with clear sourcing: capabilities, limits, risks, and practical implications for everyday people.
Field Guide for an Emerging Era
A Country of Geniuses in a Data Center
The phrase comes from Dario Amodei (Anthropic), first in Machines of Loving Grace (October 2024) and repeated in The Adolescence of Technology (January 2026). He uses it as a metaphor for millions of highly capable AI instances running in parallel, at machine speed, with real-world economic and geopolitical consequences.
“A country of geniuses in a data center.”
Dario Amodei, describing his definition of “powerful AI.” 2024 essay / 2026 essay
In his framing, “powerful AI” means systems that can:
- Outperform top human experts across most cognitive fields.
- Act in the world through software tools, communication, and decision workflows.
- Work autonomously for sustained periods without constant human steering.
Working View
AGI-like capability may already have appeared by 2024 in practice. Definitions are disputed, so this site tracks both evidence and disagreement.
Core Focus
Preparation, documentation, and human agency: helping people understand what is changing and how to respond with intent.
Evidence Markers
Why some people date AGI-like capability to 2024
- In a controlled 2024 two-player Turing test study, GPT-4 was judged human 54% of the time. A 2025 follow-up using Turing's original three-party format found GPT-4.5 with a persona prompt was judged human 73% of the time — more convincing than real humans. Source / Source
- OpenAI reported o1 at the 89th percentile in competitive programming (Codeforces) in September 2024. Source
- OpenAI reported GPT-4 scoring around the top 10% on a simulated bar exam in 2023. Source
These are capability markers, not a universally accepted AGI definition.
Grounded Timeline
Key moments that shaped modern AI
-
Dartmouth Summer Research Project
The term artificial intelligence was introduced in the Dartmouth proposal.
Source -
AlexNet wins ImageNet
Deep convolutional networks demonstrated a major leap in large-scale visual recognition.
Source -
Transformers introduced
Attention Is All You Need laid the architecture foundation for many modern language models.
Source
Community Direction
Planned structure for participation
Phase 1
Curated knowledge base
Short explainers, source links, and a shared glossary that keeps terminology precise.
Phase 2
Moderated discussion rooms
Focused conversation threads: alignment, economics, policy, labor, and technical frontiers.
Phase 3
Optional collaborations
Demand-led experiments only: lightweight projects if the community asks for them and commits to run them.
Reading List
Start here
- Machines of Loving Grace (2024)
- The Adolescence of Technology (2026)
- Attention Is All You Need (2017)
- Stanford AI Index (annual)
- NIST AI Risk Management Framework
If you want to help shape the community format, email [email protected].