Hey, in this issue: Hidden technical debt of AI systems; How to correctly use MCP servers with your AI Agents; Four ways Google Research scientists have been using Empirical Research Assistance; DeepSeek V4 compressed attention: How the KV-cache shrinks to just 2%; Breaking Opus 4.7 with ChatGPT; Agent skills for Google products and technologies; and more.
Hidden technical debt of AI systems: Agent runtime
How to correctly use MCP servers with your AI Agents
[Google] Four ways Google Research scientists have been using Empirical Research Assistance
Benchmarking multimodal document search in OpenSearch: Three approaches compared
DeepSeek V4 compressed attention: How the KV-cache shrinks to just 2%
Two years of vector search at Notion: 10x scale, 1/10th cost
[Project] Breaking Opus 4.7 with ChatGPT (hacking Claude’s memory)
This newsletter has no ads. Please support my regular work on curating this newsletter by becoming a Cool Supporter 🚀
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.