You only need about seven core building blocks to solve almost any business problem with AI. 1. Intelligence (LLM)
2. Memory
3. Tools
4. Validation
5. Control
6. Recovery
7. Feedback
PROMPT ENGINEERING
-
Seven Core Building Blocks to Solve Business Problems
By
–
-
Pre-processing Information for LLM Problem Solving
By
–
You need to pre-process all available information, prompts, and user input so the LLM can easily and reliably solve the problem. This is the most fundamental skill in working with LLMs.
-
LLM Context Engineering: Sending Right Context to Right Model
By
–
When you do make that LLM call, it's all about context engineering. To get a good answer back, you need the right context at the right time sent to the right model.
-
Strategic LLM Integration Over Full Automation Frameworks
By
–
They're mostly deterministic software with strategic LLM calls placed exactly where they add value. The problem is that most frameworks push the "give an LLM some tools and let it figure everything out" approach.
-
Building Cost-Effective AI Agents Without Expensive LLM Calls
By
–
Making an LLM API call is the most expensive and most dangerous operation in modern software development. While incredibly powerful, you want to avoid it at all costs and only use it when absolutely necessary. Here's how to build AI agents that actually work:
-
Zenith and Summit: GPT-5 Model Specialization Roles
By
–
It feels like Zenith might be the creative / normal part of GPT-5 and Summit is where the coding tasks would be routed. I haven't come across Zenith in any coding tasks, but in regular questions it tends to come up once in a while
-
Prompting AI Systems to Avoid Reinforcement Learning from Human Feedback
By
–
Gotta ask it to make stuff for you or you get RLHF'ed
-
Building a Voice-to-Diagram Note App with Kimi K2 and Groq
By
–
kimi k2 @Kimi_Moonshot + @GroqInc vibe coding one shot create a gradio note taking app that generates diagrams as you speak using Whisper + flux with @FAL + @huggingface inference providers, docs in prompt
-
AI-Assisted Programming: Beyond Vibe Coding Criticism
By
–
I've been doing the same, and writing about it online since 2022 https://
simonwillison.net/tags/ai-assist
ed-programming/
… I doubt that's what people who are blaming the Tea incident on "vibe coding" mean -
Bypassing AI Safety Features: Claude and Codex Dangerous Flags
By
–
I wrote this about my own process a few months ago, but it needs an update now I've started embracing "claude –dangerously-skip-permissions" and "codex –dangerously-bypass-approvals-and-sandbox"