My bets on open models, mid-2026
The open model ecosystem faces a capability gap compared to closed labs, with closed models currently demonstrating greater robustness and utility. Open models, while technically strong, struggle with economic constraints and evolving regulations, leading to differing trajectories in Chinese labs. The future of open models depends on real AI revenue and rapid model improvement, with the U.S. expected to regain ground in adoption metrics by 2027. The open-closed gap is influenced by economic factors, and open models are increasingly adopted in repetitive automation tasks, driven by new AI-native applications and business backend automation.
More from Interconnects
I wrote an AI textbook — how long until AI can do it better?
Reflections on AI's writing ability and how AI models get more capable.
interconnects.ai··Mads Kristian Nylund·AI EducationAI GovernanceAI Ethics
5 useful things you'll learn in my new post-training textbook (shipping now!)
After a few long years of finding time to document my lessons from training open models, my post-training book is done!
interconnects.ai··Mads Kristian Nylund·AI EducationAI DevelopmentAI Tools
Lessons from the hacks
Musings on model alignment, what determines safety, and where we go from here.
interconnects.ai··Mads Kristian Nylund·AI InfrastructureAI GovernanceAI Ethics