Jonathan Harris cuts through Model Hype, Workflow Automation, AI Governance, and AI Costs in this 60-minute Turing’s Torch: Artificial Intelligence Weekly briefing. The point is not to cheer every announcement from the pavement. It is to work out what is useful, what is undercooked, and who carries the risk once the demo glow wears off.

Expect plain-English context on power, money, data, labour and control, with the usual vendor fireworks left outside where they belong.

Jonathan Harris cuts through the AI fanfare this week, examining the practicalities of building intelligent systems. We look at the surprising value of live coding sessions for knowledge transfer and the real-world challenges of physical AI, from motors to maintenance.

Plus, the unglamorous but crucial role of data governance and context management in making AI useful, and the quiet shift towards voice as a default developer tool. It’s about what actually works, not just what sounds good in a press release.

Jonathan Harris cuts through Workflow Automation, AI Governance, and AI Costs in this 30-minute Turing’s Torch: Artificial Intelligence Weekly briefing. The point is not to cheer every announcement from the pavement. It is to work out what is useful, what is undercooked, and who carries the risk once the demo glow wears off.

Expect plain-English context on power, money, data, labour and control, with the usual vendor fireworks left outside where they belong.

A blunt, British take on what actually mattered in AI this week — and what was just noise dressed as a breakthrough.

A blunt, British take on what actually mattered in AI this week — and what was just noise dressed as a breakthrough.

AI accountability moves to boardrooms amid regulation. Can companies be transparent, or is it window dressing? Agentic AI costs exceed forecasts, raising questions about financial audits. Nvidia's AITune aims to streamline deep learning deployment. Will it democratise AI or concentrate power? Washington considers AI regulation to curb risks without stifling progress.

Social media monitoring in the Philippines aids disaster response, but misinformation is a risk. Knowledge distillation streamlines AI deployment, raising bias concerns. OpenClaw enables local AI execution with enhanced security, shifting power away from cloud providers. Alibaba's VimRAG improves AI's handling of visual data, though its long-term impact is uncertain.

OpenAI floats a four-day work week and robot taxes, sparking debate about AI's impact on labour and wealth. Is it visionary or a publicity stunt? Meanwhile, businesses grapple with redesigning processes for AI, and the need for robust AI governance frameworks intensifies amid security concerns. Plus, the latest AI systems have a memory problem. They struggle to retain context, hampering their usefulness.

This week, signal vs. noise in AI. Glia wins an award for AI safety in banking, highlighting data privacy and security. SAP partners with ANYbotics to integrate robots into industrial ops, raising automation questions. Microsoft releases multilingual text models, and Alibaba updates Qwen for integrated media. Plus, AI struggles with cultural nuances, and an AI attacks a human coder. We also examine techniques to bridge the gap between AI demonstrations and useful products.

A blunt, British take on what actually mattered in AI this week — and what was just noise dressed as a breakthrough.

A blunt, British take on what actually mattered in AI this week — and what was just noise dressed as a breakthrough.