Oct 2025
A write up of an event celebrating the 75th Anniversary of the Turing Test, noting that while modern AI can often appear human in conversation, the test’s focus on imitation may be outdated. An impressive array of speakers and panellists called for new frameworks to better assess understanding and intelligence in today’s AI systems.
Listen to NotebookLM podcast of this post:
July 2025
Many organisations face “data chaos”: siloed, duplicated, stale and ungoverned assets, plus batch-only pipelines that starve AI models of timely context. The remedy requires foundations, not another tool. With trustworthy, well-modelled and accessible data, AI moves from proofs of concept to durable, measurable business outcomes.
Listen to NotebookLM podcast version of this post:
June 2025
Tech teams can struggle to construct and align OKRs, metrics, and roadmaps if they don’t have much prior context. This post outlines some of the common challenges and introduces an app called Team Tapestry specifically designed to support teams to build a unified vision while undertaking quarterly planning.
NotebookLM podcast of this post:
April 2025
AI code generation tools are profoundly reshaping how software is built. This post reflects on my hands-on experiences with tools like Lovable, Cursor, and ChatGPT, compares design-led and engineering-led approaches, and introduces a combined workflow that combines these tools to deliver high quality working applications fast.
NotebookLM podcast of this post:
Sept 2024
At the SaaS CTO Unconference in London in September 2024, a panel explored whether AI is a lasting transformation or just hype. Drawing on tech history, I outlined four criteria for identifying transformative technology such as AI: user value, ease of integration, platform flexibility, and openness. AI’s impact will ultimately depend on collective choices we make. This post also introduces the POMIC framework and the concept of POMIC shapes. The ideas here have since evolved into the AI Future Ready framework.
NotebookLM podcast of this post: