Home›AI ResearchResearch acceleration: The jagged frontierOOpen AI News·Published Nov 18, 2026Read full storyOpenAI evaluates the impact of coding agents on accelerating AI research, noting their uneven performance and implications for recursive self-improvement in AI development.More on this storyAI agents ace 79% of training-run checks but stumble on core software engineeringNov 21· 2 sourcesOpenAI's own researchers now say they value their AI tools more than nearly half their colleaguesNov 19· The Decoder#agentic-software-development#openai