Posts
- LatestThe Attention Plane: What Makes a Voice Assistant an AssistantThe difference between chatting with an LLM and having an assistant is who manages your attention.
- Conversation ≠ intelligenceWWDC 2026 as a natural experiment: model progress does not produce conversation.
- From Push-to-Talk to Natural Conversation: How Voice Pipelines WorkFour pipeline architectures for voice AI, and how to choose the right tier for your use case.
- Voice Conversation ≠ Text Chat + AudioEight assumptions that break when voice is treated as text with an audio layer.
- Building a Gen AI ProductNotes I started writing while building BookmarksGPT.
I'm George. I use technology to solve interesting problems, and love to geek out about Generative AI and guitars. I've worked across many industries, as both an engineer and a manager.
Publications
- Active Idle Screen 2009–2011: Who Will Own the ScreenApple Books
- Activating the Idle Screen: Uncharted TerritoryInforma Telecoms & Media
- Techniques for Content-Based Image Characterization in Wavelets DomainPhD Thesis, University of South Wales
- Texture-based Image Retrieval in Wavelets Compressed DomainIEEE International Conference on Image Processing
- Faster Distance Calculation in Image Retrieval SystemsIEEE International Conference on Consumer Electronics
- Quadtree Based Image Indexing in Wavelets Compressed DomainEurographics UK Conference, IEEE Computer Society Press
Speaking & Teaching
I've spoken at events including the World Bank Global Forum on Innovation & Technology, Future Internet Assembly, and Smart Cities Industry Summit, and have been an invited expert at Oxford University, guest lecturer at UniCT, and startup mentor at LAUNCHub and Startup Weekend.