Tag: ai

From Code Generation to AI-Native: Insights from the Front Lines

An analysis of a panel discussion with Anthropic, Trae, and Replit, connecting their best practices for AI-native development to the evolving role of the software engineer.

Revisiting Vibe Coding: An Analysis of AWS's Practical Guide

An analysis of the AWS guide on vibe coding, comparing its structured, tool-centric approach with earlier conclusions on agentic workflows and human-led architecture.

A Practical Guide to Coding with LLMs: Dos and Don'ts

A structured approach to using AI coding assistants, moving from contemplative prompting to scalable agentic workflows.

Is AI the Next Microprocessor or the Next Shipping Container?

An analysis of whether AI will create new wealth like the microprocessor or primarily benefit incumbents and customers like shipping containerisation.

Working with AI Wizards and the Need for Explainability

Ethan Mollick's 'wizard' analogy for AI highlights a growing problem: as models become more powerful, their processes become more opaque, creating a critical need for explainability.

Contrasting Memory Philosophies: Claude's Explicit Tools vs. ChatGPT's Automatic Profiles

A look at the different approaches to memory in Claude and ChatGPT, highlighting the trade-offs between transparency and convenience.

A Structured Methodology for Disciplined AI Software Development

A summary of the 'Disciplined AI Software Development' methodology, which provides a framework for managing AI collaboration in coding projects.

Where's the Shovelware? The AI Coding Claims Conundrum

An analysis of why AI coding tools haven't led to a surge in software production, despite widespread adoption and bold productivity claims.

Building Reliable AI Agents: Notes on Evaluation and Contracts

A concise look at why evaluating AI agents is essential and how the 'contractor' model can improve their reliability in business.

Quantifying the Environmental Impact of Large Language Models

Mistral AI has published a comprehensive lifecycle analysis of its models, setting a potential standard for measuring the environmental impact of artificial intelligence.

Entering the probabilistic era of software

A look at Gian Segato's essay on the shift from deterministic to probabilistic software development and its implications for business.

The GenAI divide report: Scrutiny, hype, and the reality of AI adoption

An analysis of the controversial MIT NANDA report on AI ROI, contrasting its claims with critical perspectives and exploring the real challenges of enterprise AI adoption.

GitHub's playbook for an AI-powered workforce

A summary of GitHub's internal playbook for scaling AI adoption, focusing on change management and organisational strategy.

MCP servers: less is more

An analysis of the risks associated with Model Context Protocol (MCP) servers, focusing on token consumption and security vulnerabilities.

What are the most profitable uses of ChatGPT?

A summary of real-world examples from Reddit showing how people use AI tools to save money, improve their work, and solve personal problems.

Fixing the context window for LLM agents

A look at key strategies for managing context in LLM agents, inspired by an article from dbreunig.com and insights from Cognition AI.

Rethinking RAG with visual document analysis

An interesting approach from Morphik that uses images of pages for Retrieval-Augmented Generation, avoiding traditional parsing issues.

Onboarding AI with READMEs and quality gates

A practical approach to structuring project documentation using READMEs for context and automated checks as 'Quality Gates' to improve AI-assisted development.

Simon Willison on the lethal trifecta and MCP security

A look at Simon Willison's latest talk on AI security, focusing on his 'lethal trifecta' concept and the risks of the Model Context Protocol (MCP).

A deeper look at AI's impact through the SPACE framework

A reflection on a study that uses the SPACE framework to measure the real-world impact of AI on developer productivity, satisfaction, and collaboration.

Experimenting with Kilo Code and the Model Context Protocol

A look at my first experiences with Kilo Code, an open-source alternative to Cursor, and the process of getting the Playwright MCP working.

The bitter lesson for organisations

Ethan Mollick's article on 'The Bitter Lesson' suggests that brute-force AI may outperform human-designed processes, but where do humans still hold the advantage?

A refreshing, back-to-basics approach to coding with AI

A look at antirez's argument for using LLMs as direct, human-controlled add-ons, avoiding agents and retaining full control over the development process.

Has self-service BI finally arrived with AI?

A look at how conversational AI and Model Context Protocol (MCP) could revolutionise business intelligence, with a dose of healthy scepticism.

Frameworks for measuring developer productivity

A look at the DX frameworks for measuring AI and developer productivity, placed in the context of other models like DORA and SPACE.

The state of software engineering with LLMs in 2025

A look at how LLMs are changing software development, inspired by Gergely Orosz's recent article, and my own plans for adopting these new tools.

A practical LLM coding workflow and a reflection on teamwork

A summary of Harper's LLM codegen workflow and how structured rulebooks might solve the 'solo developer' problem he identifies.

How to use AI without damaging your thinking

Reflections on Ethan Mollick's article about using AI to enhance, not hinder, our thinking, with practical tips for learning and writing.

From theory to practice: How organisations are adopting AI

A look at a Stack Overflow article showing real-world examples of AI implementation at companies like Cloudflare, GitHub, and Abnormal AI.

Reflecting on AI's real impact on engineering leadership

A look at the 2025 LeadDev report on AI in engineering, and why a clear roadmap is more important than ever for successful adoption.

The last hurrah of human coding

A reflection on Alex MacCaw's concept of 'vibe coding' for senior engineers and how it aligns with the need for structured, architect-led AI collaboration.

Using an LLM as a personal tutor for my systems engineering course

A detailed look at my workflow for using a custom-prompted LLM to create effective study materials for a university course.

Beyond the vibe: structuring AI-assisted development

Discover how to move from chaotic 'vibe coding' to a structured workflow by applying advanced techniques for steering AI collaboration tools like Claude, Cline, and Cursor.

Security risks in LLM agents: Simon Willison’s insights

A summary of Simon Willison’s recent posts on security risks in LLM agents, including the lethal trifecta, design patterns, and real-world vulnerabilities.

Understanding MCP: A New Standard for AI Integration

Exploring Anthropic's Model Context Protocol and its potential to transform how AI interacts with software tools.

Agents: Programming with feedback-driven LLMs

How LLM agents become dramatically more capable when given tools to interact with their environment.

Building a modern resume: from manual tweaks to AI-powered automation

Discover how I transitioned from tedious manual resume updates to an efficient, AI-assisted workflow using Reactive-Resume and a custom JSON-to-Markdown converter.

Emerging developer patterns for the AI era

Overview of nine patterns for software development with AI agents from a16z.

Making AI Work: Leadership, Lab, and Crowd

Ethan Mollick discusses the organizational changes and challenges companies face when integrating AI, focusing on leadership, experimentation, and learning.

More tags: