Seeing research and business perspectives side by side was especially interesting. Researchers highlighted the difficulties of accuracy, control and integration into workflows. Businesses showed how AI is already being used beyond experiments.
Hallucinations and uncertainty are not simply random noise. They can be better understood, measured and accounted for through checks in a workflow.
Adopting AI is not a matter of flipping a switch. People retain responsibility, and specialist knowledge becomes more important, not less.
Useful discussion of AI rests on a method, measurement and an honest understanding of limitations, rather than slogans.
There is a substantial need for basic knowledge. People need more than a list of tools: they need a clear understanding of where to start and what to avoid.