Blog, keynotes, publications and recordings
Highly effective knowledge management is no longer the privilege of elite companies.
I recall visiting a McKinsey office maybe 15 years ago and being very impressed by their structured approach to organising knowledge.
from RAGs to riches
Retrieval Augmented Generation is currently the best way to mitigate Large Language Model hallucinations.
Guilty Pleasure – comparing DALL-E, MidJourney, Stable Diffusion and Photoshop
AI-based image generation tools are transforming the way we perceive and create art. I explored four such tools to understand their unique capabilities and limitations. The test subject was the dame blanche website (and Instagram feed) that I have finally released...
This is not my first hype cycle
Every time a new technology comes along, we seem to first believe that it will (a) replace all previous technologies and (b) that it will solve all future problems. Inevitably we then discover the limitations and then finally we put the new tool where it belongs, in its own niche where it shines.
Massive AI technical debt incoming…
we are well aware of model drift, where the input variables (the real world), over time, become different from the data distribution the model was trained on. It’s not so much the model that drifts here but rather the input data.
With LLMs the models themselves also ‘drift’ in that
Why are manhole covers round?
When typing this question into ChatGPT I was impressed by the response that suggests deep insight into the relationship between the shape, weight and strength of a manhole cover, the challenges to manufacture and manipulate it as well as worker safety concerns. Surely an AI needs a model of the world and true reasoning capability to arrive at this answer?