Google DeepMind on ai-engineering
Google DeepMindThis talk explores the rapidly evolving field of agentic AI, focusing on the tension between AI-driven speed and the need for human-centric quality and control. Key themes include the shift in software engineering bottlenecks from intelligence to human attention, the emergence of faster AI models, a
AI JasonThe latest AI models, particularly since December 2025, have achieved a "step function improvement" allowing for fully autonomous, long-running tasks. This shifts AI from co-pilot systems to continuous, independent agents. The key to successful deployment lies in "harness engineering," focusing on c
Crux: Person A emphasizes ongoing human-AI collaboration and human-centric control, while Person B suggests AI models are capable of fully autonomous, long-running tasks, shifting away from co-pilot systems.