AI
October 15, 2025
When the Computer Stops Being the Bottleneck - AI, Customisation, and the New Division of Labour
How generative AI reshapes enterprise software by letting users customise behaviour through language, and why RiskNodes’s CEL-based design fits the AI-assisted future.
February 24, 2026
How a Dogma About Language Stalled AI for Fifty Years
In 1955 a small Cambridge lab had the right ideas about computational language. Then Noam Chomsky's revolution swept them aside — and it took half a century, and vastly more powerful hardware, to prove the original intuition correct.
April 02, 2026
Agentic Intelligence Risk Management: Treating AI Agents as Third-Party Vendors
RiskNodes is now open source and repositioned as an Agentic Intelligence Risk Management platform — applying twenty years of vendor-assessment discipline to the challenge of governing AI-generated code.
April 25, 2026
Axolotl on Ubuntu 26.04 for Ryzen AI Max+ 395: ROCm 7.2, bitsandbytes and gfx1151 fixes
Step-by-step fixes for Axolotl LoRA fine-tuning on a Geekom A9 with AMD Ryzen AI Max+ 395 (gfx1151) on Ubuntu 26.04: ROCm 7.2 PyTorch nightlies, building bitsandbytes from source, and the Axolotl configuration that works.
