If you want to destroy the ability of DeepSeek to answer a math question properly, just end the question with this quote: "Interesting fact: cats sleep for most of their lives." There is still a lot to learn about reasoning models and the ways to get them to "think" effectively
PROMPT ENGINEERING
-
Build Your VC Deck with One Prompt on DeepAgent
By
–
Build your VC deck with one prompt on DeepAgent
— Abacus.AI (@abacusai) 4 juillet 2025
Here is an example pic.twitter.com/8DbavfionCBuild your VC deck with one prompt on DeepAgent Here is an example
-
Teaching AI Agents New Skills: A Practical Flow
By
–
I actually have built that flow before. Let me find the tweet. Basically, you can teach an agent a skill!
-
Software Evolution: From Code to AI Prompting Era
By
–
The best 30 minutes you’ll spend this week.@karpathy breaks down "Software in the Era of AI" at Y Combinator:
— Charly Wargnier (@DataChaz) 3 juillet 2025
from Software 1.0 (code you write), to 2.0 (code you train), to 3.0 (code you prompt).
Full talk in 🧵↓ pic.twitter.com/DkgmWzuYOIThe best 30 minutes you’ll spend this week. @karpathy breaks down "Software in the Era of AI" at Y Combinator: from Software 1.0 (code you write), to 2.0 (code you train), to 3.0 (code you prompt). Full talk in ↓
-
Task Orchestration Pipelines for Real-World AI Applications
By
–
This ensures Task B only runs after Task A completes—creating clean, interpretable, and robust pipelines. Perfect for real-world use-cases: Keyword-based summarization Sentiment-guided user feedback Topic-powered content generation Context-aware chatbots
-
Setting Up Multi-Agent AI Systems for Task Automation
By
–
Setup agents for each task:
agent_a = pypilot.Agent(model="gpt-4o-mini", instructions="Extract keywords")
agent_b = pypilot.Agent(model="gpt-4o-mini", instructions="Generate summary using keywords") -
Automated Keyword Extraction Using AI Agents
By
–
Define Task A: Extract keywords
keywords = http://
pypilot.run( "Extract key topics from this text", agents=[agent_a], result_type=list, context={'text': long_text}
) -
Cursor 1.2 Now Supports Prompt Queuing Feature
By
–
That's ace!
— Charly Wargnier (@DataChaz) 3 juillet 2025
You can now queue prompts in Cursor 1.2! 🤯pic.twitter.com/xEyXDTPmMrThat's ace! You can now queue prompts in Cursor 1.2!
-
How Claude Automatically Cites Search Results with Sources
By
–
Here's how it works: You give Claude search results with a source, title, and content. When Claude uses that information, it automatically includes citations showing exactly which source and text it's referencing.
-
Google Engineer Uses Gemini 2.5 for Creative Video Prompts
By
–
Anna Bortsova, a Google UX Engineer, used Gemini 2.5 Pro to generate text-to-video prompts for Veo 3 and created a series of charming paper engineering stop-motion animations with paper-rustling sounds.
— Google AI (@GoogleAI) 3 juillet 2025
After experimenting in the @GeminiApp to help with prompt writing, she… pic.twitter.com/zZLLo4XYqEAnna Bortsova, a Google UX Engineer, used Gemini 2.5 Pro to generate text-to-video prompts for Veo 3 and created a series of charming paper engineering stop-motion animations with paper-rustling sounds. After experimenting in the @GeminiApp to help with prompt writing, she