machine-learning.
17 writings found
Page 2
Teaching AI Models to Actually Read Maps: Google's MapTrace Pipeline
Google researchers built a synthetic data pipeline to teach multimodal LLMs spatial reasoning. Turns out, tracing paths on maps is surprisingly hard for AI.
The Multi-Agent Myth: Why More AI Agents Aren't Always Better
Google Research reveals the first quantitative scaling principles for AI agents, showing when multi-agent systems help and when they catastrophically fail.