Panel Discussion: Building Trust in AI Predictions by Validation & Benchmarking
4:00 pm - Wednesday 13th January 2027
- Establish rigorous validation frameworks that identify the 50% of AI outputs that are hallucinations, ensuring your teams trust and adopt computational tools rather than dismissing them due to poor early experiences
- Learn how to benchmark AI predictions against experimental data, understanding which use cases deliver value and which remain aspirational
- Discover strategies to educate scientific teams on AI capabilities and limitations, setting realistic expectations and driving productive adoption while avoiding the “magic bullet” trap
