By Maxwell Zeff
Publication Date: 2026-05-27 14:00:00
A group of AI researchers who previously worked at Google DeepMind, Apple, OpenAI and Meta Superintelligence Labs announced Wednesday that they are launching a new startup called Trajectory that aims to help companies regularly improve their AI products through training on real-world user interactions.
Trajectory aims to build a platform for AI that can continuously learn – a capability that researchers have long viewed as a major obstacle to further AI advances. OpenAI, Google, and Anthropic are finding success in training increasingly powerful versions of AI models, particularly for areas like coding, math, and science. However, these systems no longer become intelligent once they have completed their training. While there have been some recent breakthroughs in continuous learning, tech companies generally struggle to develop AI products that learn from their mistakes in real time. In December 2025, Turing Prize winner Richard Sutton argued at NeurIPS, one of the largest annual AI research conferences…
