Maximilian Schreiner

4 days ago | the-decoder.com | Maximilian Schreiner

Summary A new study from Stanford University finds that AI agents can get much better at solving complex tasks simply by learning from their own successful experiences. So far, building effective AI agents has often meant a lot of manual effort: fine-tuned prompts, handpicked sample sets, or specialized action spaces. These methods work, but they are time-consuming and hard to scale.

5 days ago | the-decoder.com | Maximilian Schreiner

Summary Retrieval-augmented generation (RAG) promises to help medical AI systems deliver up-to-date and reliable answers. But a new review shows that, so far, RAG rarely works as intended in real-world healthcare settings—and technical, regulatory, and infrastructure hurdles are slowing its adoption. Large language models have proven powerful across a range of fields and are already in use in many industries. Medicine, however, is a different story.

6 days ago | the-decoder.com | Maximilian Schreiner

Summary Alibaba’s research lab Tongyi has introduced ZeroSearch, a new method for training large language models to handle search tasks—without relying on real web searches. For chatbots to answer questions accurately, especially when their built-in knowledge isn’t enough, they need to learn how to find information on the fly. Most current approaches use reinforcement learning (RL) and depend on actual search engines like Google to teach this skill.

6 days ago | the-decoder.com | Maximilian Schreiner

Google is now using AI models to protect Chrome users from online scams. On desktop, the company has rolled out its local Gemini Nano language model to quickly spot fraudulent websites, including ones that have never been seen before. On Android, Chrome will now warn users about suspicious notifications sent by websites. Google says these changes are part of a broader effort to improve security, which also includes the "Enhanced Protection" feature in Safe Browsing.

1 week ago | the-decoder.com | Maximilian Schreiner

Netflix is rolling out a new search feature powered by OpenAI's ChatGPT technology, allowing users to look for content using natural language. Instead of typing in titles or keywords, people can now search with phrases like "I want something funny, but not silly." The new search is launching as a beta on iOS devices, with early tests already taking place in Australia and New Zealand, according to Bloomberg.