Posts

#4636 Moonshot AI Projects; +First Principles to Design the First Computer

-
#4636 Moonshot AI Projects; +First Principles to Design the First Computer Y Combinator: Jeff Dean: The 1% Rule for Building in AI https://youtu.be/CxXgV54KzpQ AI Is Improving Faster Than Ever | MOONSHOTS https://youtu.be/-LzH932gIJo #4636  Moonshot AI projects Depending on the context, "Moonshot AI" usually refers to either the prominent Beijing-based AI research lab (creators of Kimi ), the open-source Project Moonshot evaluation toolkit, or the concept of high-risk, high-reward "moonshot" initiatives in artificial intelligence. Here is a breakdown of the primary projects associated with each: 1. Moonshot AI (The Frontier AI Research Lab) Founded by Yang Zhilin, Beijing-based Moonshot AI is one of China's primary "AI Tigers," known for pushing the boundaries of ultra-long context windows, multimodal agentic systems, and custom model architectures. Core Models & Architectures Kimi K3: A 2.8-trillion parameter open model featuring a 1-million-t...
Post a Comment
Read more

#4635 Critique Agent Update to Other AI Agent System Instructions

-
#4635 Critique Agent Update to Other AI Agent System Instructions #4635  Given that AI Agents have system instructions, can a follow-up critical agent suggest changes to the previous agent's instructions in an attempt to improve its response? Yes, absolutely. In multi-agent orchestration architectures, having an Evaluator/Critic Agent dynamically modify or refine the System Instructions (Prompts) of a Worker/Generator Agent is a core design pattern for iterative improvement. This approach goes beyond simple output feedback (e.g., "Rewrite this to be shorter" ) by addressing the root cause of an inadequate output: the operational framing and rules governing the generating agent. Here is how this dynamic works in practice, along with key design considerations: 1. The Meta-Prompting / Prompt Optimization Loop Rather than having the Critic directly fix the Generator's final response, the workflow operates in a meta-feedback loop: Generation: Agent A (Worker) receives...
Post a Comment
Read more

#4634 PDF -> AI-Gen'd Knowledge Graph with Embedding

-
#4634 PDF -> AI-Gen'd Knowledge Graph with Embedding https://learningfromchatgpt.blogspot.com/2026/07/4620-knowledge-graph-classical.html #4634  Can Gemini 3.6 read a PDF and generate complete correct text from a specified section of the document? Yes , Gemini 3.6 can process PDF files and extract exact text from a designated section, but there are a few nuances regarding how it handles the task depending on the document type: 1. How Gemini 3.6 Handles PDF Inputs Gemini processes PDFs natively through multimodal computer vision and layout analysis rather than relying solely on traditional text-scraping OCR. Digital (Text-Based) PDFs: Near-100% accuracy. If the text is natively embedded in the PDF, Gemini can read, isolate, and reproduce a specified section exactly as written. Scanned / Image-Based PDFs: Highly accurate, but subject to OCR visual limitations. If the scan is clean and high-resolution, it transcribes text with strong fidelity. If the source material features...
Post a Comment
Read more