According to @DeepLearningAI, researchers from Mila, Microsoft, and academic partners proposed Delethink, a reinforcement learning method that trains language models to periodically truncate their chain-of-thought to manage long-context reasoning more efficiently, source: @DeepLearningAI, Twitter, Jan 17, 2026. The post adds that Delethink reduces the cost of long-context reasoning and improves performance, a combination that directly highlights lower inference cost and higher throughput for LLM workflows, source: @DeepLearningAI, Twitter, Jan 17, 2026.

According to @DeepLearningAI, Andrew Ng introduced Build with Andrew, a no-code course that teaches beginners to describe an app idea in plain language and use AI to build a working, shareable web application. According to @DeepLearningAI, OpenAI fine-tuned GPT-5 Thinking to admit when it breaks rules. According to @DeepLearningAI, Shanghai AI Lab released SCP, an open protocol for scientific agents. According to @DeepLearningAI, a Microsoft study finds chatbot use shifts by device, time of day, and season. According to @DeepLearningAI, researchers introduced Delethink, a method for cost-efficient long-context reasoning. According to @DeepLearningAI, these items are included in this week’s The Batch newsletter update.