This one particularly leverages the long context window of llms!
PROMPT ENGINEERING
-
Using Role-Based Prompting for Improved LLM Outputs
By
–
“Use role-based instructions” changed the game for me. Just tried “You’re a VC writing a funding email”—output was .
-
AI Coding Capabilities Enable Complex Web Apps with Fewer Prompts
By
–
With updated coding and reasoning capabilities, it’s especially strong at building interactive web apps. What used to take 50+ prompts a year ago with 1.5 Flash (left) now takes <10 with far better quality and much more complex UI (right). pic.twitter.com/QoIqScd43U
— Google AI (@GoogleAI) 6 mai 2025With updated coding and reasoning capabilities, it’s especially strong at building interactive web apps. What used to take 50+ prompts a year ago with 1.5 Flash (left) now takes <10 with far better quality and much more complex UI (right).
-
Using o3 Prompts to Configure ChatGPT as a Business Consultant
By
–
Steal my o3 prompt to transform ChatGPT into your $50K business growth consultant. ———————————-
HYPERGROWTH CONSULTANT
———————————- HYPERGROWTH ARCHITECT SYSTEM
You are an elite growth strategist who has led companies from zero to -
Best Course to Learn Up-to-Date RAG Techniques
By
–
“Best course I found to learn up-to-date RAG techniques.” — Patrick Drechsler (Data Scientist)
-
Comprehensive LLM Engineering Guide Praised for Clarity
By
–
“Amazed at the thoroughness and clarity…covers the whole spectrum of LLM engineering end-to-end.” — Carlo Casorzo (AI Engineer)
-
AI Models Ignoring Word Count Constraints in Responses
By
–
I also wish it didn't think every request has to be 15 pages in length. I specifically ask for <1500 words (I want it to tackle hard problems but don't need a dissertation response) and it can't fathom that I don't want 150,000 words.
-
AI System Leaks Information Outside Project Scope
By
–
It still seems to pull info from convos outside the group (project). Even if you tell it not to.
-
ContextGem Leverages LLM Long Context Windows for Extraction
By
–
ContextGem particularly leverages the long context window of LLMs for extraction tasks!
-
Gemini 2.5 Pro enables complex web apps from single prompt
By
–
With this update, you can create even more complex web apps from a single prompt. See a side by side (new vs old 2.5 Pro) on building interactive learning apps from YouTube videos.
— Oriol Vinyals (@OriolVinyalsML) 6 mai 2025
Try the new Gemini 2.5 Pro in AI Studio, Vertex AI, Gemini app & Canvas. https://t.co/DAUx06ITAw pic.twitter.com/SXsOirM1d3With this update, you can create even more complex web apps from a single prompt. See a side by side (new vs old 2.5 Pro) on building interactive learning apps from YouTube videos. Try the new Gemini 2.5 Pro in AI Studio, Vertex AI, Gemini app & Canvas. http://
aistudio.google.com