Don Kim

In "Research + draft" step: A. "Research" actually produces an AI-generated summary in a structured format, which is passed to the "draft" node B. "Draft" creates an initial post, the prompt is injected with voice_profile.md file, which is supposed to mimic my personal style. I am thinking to try to optimize voice_profile.md first. Any other suggestions?

@Dondyu the scan → draft → polish flow is collapsing the thinking and the writing into the same step. the human voice usually lives in the gap between those two. what does your research step actually hand to the draft step?

Built a 3-stage AI pipeline to amplify my own voice: scan HN + Substack → research + draft → polish + fact-check. Ran the entire pipeline with minimal human modification, and the content still reads very AI. Going to experiment with prompts and workflow arch to improve it.

@omarsar0 In enterprise context, domain knowledge is usually stored in the minds of senior team members. I expect LLM Wikis should be a way to democratize that knowledge systematically. Not only for LLMs, but also for humans. Gonna give it a try.

Releasing a little skill to help build LLM Wikis. More people should be building LLM wikis. I hope the skill is useful for your agents. Skill, examples, blog, and even a video tutorial included. Have fun.

@gabrielchua @aiDotEngineer @thsottiaux see you at the events!

@Dondyu @aiDotEngineer @thsottiaux thanks for catching!

Hi Singapore 🇸🇬, meet Codex 🩵 With Codex, ANYONE can build and create. We’re turning that energy up this May. We’re a diamond sponsor 💎 at @aiDotEngineer Singapore, at a bunch of events, and hosting sessions for builders at every level. Plus, @thsottiaux will be in town. There's something for everyone: > Intro workshops if you’re just getting started. > Hack nights if you want to go deep & tokenmax > Demo nights if you want to see what people are building. May is going to be fun, and we're just getting started. +++++++++++++++++ 3 May: Emergency Codex Hackathon luma.com/zprbhtzyTogeth… with @brianchew 4 May: Codex Hack Night - Design & Frontend luma.com/bw3rpffqTogeth… with @cleondesigns 5 May: Codex Demo Night - GPT-5.5 & ImageGen 2 luma.com/tu1yajd5Togeth… with @Lorong_AI and @yongquanYQ 7 May: AI Engineer UnConference luma.com/3binn6ce No talk there, but I’ll be hanging out. Come find us. Together with the @65labslah crew 9 May: AI Engineer Singapore Hackathon Link is TBA Together with the @65labslah crew 9 May: Codex workshop at The Good Hack lnkd.in/gYrUMj3g Thanks for inviting us GoodHub SEA (Joanne Tan), Open Government Products 11 May: Codex Hack Night - Computer Use luma.com/pbk5bb32 Together with @Lorong_AI and @yongquanYQ 13 May: Codex Hack Night - TBC 👀 luma.com/4v97ndhxTogeth… Together with @Lorong_AI TBA: Intro to Codex for Everyday Work - Workshop 15 May: AI Engineer Singapore - Codex Workshop 16-17 May: AI Engineer Singapore - Keynotes + Codex Booths and more ...

Setup: 1. Install the Claude browser extension + Claude Desktop 2. Open your article in the browser 3. Grant the Claude extension access to the tab 4. Type the prompt below into Claude Desktop using Haiku model Prompt: Prepare the article summary: 1. Scan the page I have open in current browser tab 2. Output the final guide as styled inline HTML. Use Visualize plugin to render HTML in this conversation directly. 3. Include all images/diagrams in the explanation. 4. At the end of the summary explicitly acknowledge 2 things: 1. Which key content points is added in the summary that wasn't in the original article 2. Which key content points from in the article have been excluded from the summary? 5. Format: - Emoji headers (🎯 What is X?, 📷 Example, etc.) - Gray concept boxes for main content - Blue callout boxes for insights - Green callout boxes for tips - Red callout boxes for warnings - 2-column comparison grids - Monospace code sections with blue left border 6. Do not append any other information other than the produced HTML guide

I'm a slow reader. Long articles destroy my focus halfway through. So I made a Claude prompt that reformats any article into a visual study guide — color-coded boxes, emoji headers, comparison grids. It only takes 10 seconds, but completely changes how I read. See before & after in images below. Prompt + setup in the replies ↓

2.While I am not familiar with detailed PCI specs, I reckon the PCI standards will need to evolve to accommodate the growing number of Voice Agentic usecases.

sierra.ai/blog/payments This is impressive for voice interfaces and made me think of 2 points: 1. Voice AI architectures cant fully outsource PCI requirements to external PSPs. Thus having PCI compliant in-house architecture is a moat, and most platforms havent caught up yet

Been thinking about a trend - as a consequence of the AI wave, the number of non-corporate people increases (e.g. soloprenuers). It may be a growing untapped market, so wondering how many underserved opportunities we may find there in the next few years.

Contributing OSS: 1. Watch GitHub repositories you are using frequently 2. See "All Activity" on github.com 3. Find issues labeled as "Help Wanted" or "Good First Issues" 4. Leave a comment like "I'll work on this" on the issue if you could do this 5. Work on it!

"Either you do product or sales. Everything else is optional"

Perspective matters more than knowledge. Prioritize searching for your own right path over pure knowledge building (but do not disregard it, however)

50 demos closer :)

Fundamentally, Apple is no longer at the forefront of innovation, but in the middle run, it will continue to deliver ROI. Credit: @BenedictEvans

I believe loyalty is the number one basis for any deep relationship twitter.com/punitsoni/stat…

A child’s talents often take a lifetime to uncover. 👇

Startup meetings be like

A thorough read of this article alone is worth about 10 credits in an undergrad Management degree #weekendread bloomberg.com/news/features/…

Not to claim that Google will not figure it out, if that happens to be true