AI Won't Replace You. Someone Who Decides Will.
The conventional AI-and-jobs conversation is stuck on the wrong question. AI can't practise intentionality — it resolves statistical patterns. That makes the human behind the prompt the entire moral equation.
What 7,308 Agent Runs Taught Me About Writing Better Skills
SkillsBench tested 84 tasks across 7,308 agent runs and found curated skills boost performance by 16pp — but 19% of tasks got worse. It is a great backdrop to the things I have learnt too.
Lawyers Got Prompt Engineering Wrong (And Why That Matters)
At TechLawFest 2025, Singapore lawyers packed a workshop on prompt engineering. Meanwhile, the technology shifted: agent skills became available. This isn't about better prompts—it's about who controls the decision-making logic. From 3-page prompts to reusable systems.
The Solo Counsel Reality: What MinLaw's AI Guidelines Miss About In-House Practice
Singapore's AI guidelines assume enterprise resources and committees. But what about solo counsel making real AI decisions with limited budgets and shadow tools? This consultation misses practitioners who need guidance most.
Introducing: Prompt Engineering for Lawyers
I introduce my latest work - a set of tutorials on prompt engineering for the legal domain and talk about my vision and aspirations for it.
Building SG Law Cookies: Insights and Reflections
I share my experience and process of generating daily newsletters from Singapore Law Watch using ChatGPT, serverless functions and web technologies.
Augmenting ChatGPT with Your Data Can Improve Its Performance
I discuss how ChatGPT can be augmented with custom data to improve its performance using Singapore law as an example.
Let ChatGPT take over your resume
I explain how to use ChatGPT to write resumes, including finding a template, directing the model to produce bullet points, and summarizing experiences. The article highlights the benefits of using ChatGPT and provides tips for using the tool effectively.
One does not simply ChatGPT a contract
Very little is understood about how ChatGPT and Large language Models could work, but that has not stopped others from having an opinion about it.
Introducing: zeekerscrapers
The latest updates to my slow-going personal project, zeeker.
