Ethan Mollick's new book, "Co-Existence," highlights the rapid advancement of AI over the past 30 months, contrasting it with his previous book, "Co-Intelligence." He notes that AI's capabilities have progressed to a point where tasks that once required extensive human-AI interaction can now be delegated to autonomous agents, significantly speeding up complex processes like research and data visualization. Mollick points to the METR Long Horizon Benchmark, which showed AI capabilities doubling eight times since GPT-4's release, and a dramatic decrease in the cost of obtaining accurate answers on benchmarks like the Google Proof Question and Answer Exam. AI
IMPACT Demonstrates how agentic AI and falling costs are reshaping human-AI collaboration and accelerating development cycles.
RANK_REASON Blog post by an influential AI commentator discussing AI progress and capabilities.
Read on One Useful Thing (Ethan Mollick) →
- Anthropic
- ChatGPT
- Claude Projects
- Co-Existence
- Co-Intelligence
- Google Proof Question and Answer Exam
- GPT-4
- METR Long Horizon Benchmark
- OpenAI
- Opus 5.5
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →