- Pointer
- Posts
- Issue #724
Issue #724
Essential Reading For Engineering Leaders
Friday 12th June’s issue is presented by Gauntlet
In this Night School, Ash Tilawat breaks down why AI-native development is moving toward a document-driven workflow. He'll show how specs act like contracts with AI agents, why planning is becoming the most important part of the build, and how decision logs create a shared constitution for architecture and product choices.
What You'll Learn:
→ How the software development lifecycle is changing inside AI-native teams
→ Why planning is becoming one of the most important bottlenecks in AI-assisted development
→ How to use specs as clear contracts for coding agents
→ How decision logs track architecture choices and guide future work
→ What separates a true AI engineer from someone who is just vibe-coding
— Wes Kao
tl;dr: Luckily, you can build the muscle of speaking up and sharing your point of view. Here are 6 principles that helped me: (1) The more controversial the idea, the higher the burden of proof. (2) Update your assumptions about how you add value. (3) Share where your hunch is coming from because it’s coming from somewhere. (4) Describe why the problem matters, so people understand why you’re speaking up. (5) Don’t rely on your credentials. Your idea should make sense on its own. (6) Use language that accurately reflects your level of certainty.
CareerGrowth Communications Tips
— Lizzie Matusov
tl;dr: “Most discussion of AI coding tools centers on how they change an individual’s output, but they are also quietly rewiring how people on a team turn to one another. This week we ask: how does GenAI change when and how members of a software development team interact?”
Management TeamHealth AI
tl;dr: In this Night School, Ash Tilawat breaks down why AI-native development is moving toward a document-driven workflow. He'll show how specs act like contracts with AI agents, why planning is becoming the most important part of the build, and how decision logs create a shared constitution for architecture and product choices. What You'll Learn: (1) How the software development lifecycle is changing inside AI-native teams. (2) Why planning is becoming one of the most important bottlenecks in AI-assisted development. (3) How to use specs as clear contracts for coding agents. (4) How decision logs track architecture choices and guide future work. (5) What separates a true AI engineer from someone who is just vibe-coding.
Promoted by Gauntlet
Event AI Agents
— Murat Demirbas
tl;dr: “The core of the paper rests on this monotonic decay argument. The sheer task-level velocity gains we see from AI coding tools start to bleed out as they move up the production hierarchy.”
Management DevEx AI
Editorial Note
I’m looking for software engineers who are currently managing or leading a team to answer one question over email.
If you are interested, click reply.
— Steve Huynh
tl;dr: “You don’t need another productivity system, or more hours in the day. All you need is to choose to do one thing at a time, to be fully in it, especially when something in you is itching to be somewhere else.” Steve provides actionable things you can do to reach this goal.
Tips Productivity Wellbeing
tl;dr: Lore saves your team's Claude Code, Codex, and Cowork conversations in one place, so you don't lose all that work. Get shareable URLs for sessions showing what worked (or didn’t work) across prompts, skills, and tool calls. Search decisions, related work, and context across your team without waiting for someone to reply.
Promoted by Tanagram
Tools+Setup KnowledgeManagement AI
— Benedict Evans
tl;dr: “Many people would like to analyse which jobs, companies and industries are most exposed to AI, and assign scores, build charts, and map that against the progress of LLMs. I think this is mostly impossible: you don’t know how the jobs will change, you don’t know what else will change around this, and you can’t measure work like that anyway.”
IndustryTrends AI
tl;dr: “Most modern LLMs share the same transformer-family skeleton. The differences come from what each one was trained on, the scale and configuration choices, and the post-training done on top. By the end, you should be able to read many modern LLM papers or model cards and know which piece of the architecture each section is talking about.”
DeepDive AI
— Radion Khait
tl;dr: “A test passes. Great! But does it really mean your code is working as expected? Not necessarily. Sometimes the values you choose in your tests can create a false sense of security, especially when dealing with default values.”
Testing CodeQuality
Null Pointer
Model Diplomacy
Hand-drawn by Manu. View the Null Pointer series.
Most Popular From Last Issue
Notable Links
Container: Linux containers using virtual machines on Mac.
Cosmos: Open platform to build Physical AI.
Headroom: Context compression tool for LLMs.
Hivemind: One brain for your agents.
SkillSpector: Security scanner for agent skills.
How did you like this issue of Pointer?1 = Didn't enjoy it all // 5 = Really enjoyed it |