Every year, Gartner’s Hype Cycle tells an interesting story — and this year’s 2025 AI Hype Cycle feels spot on for drug discovery and development

A few years ago, everyone was talking about Generative AI as if it would change everything overnight.
Now? It’s sliding down the trough of disillusionment.
People are realizing how hard it is to control hallucinations, especially when accuracy matters, like in biomedical research or pharmacovigilance.

At the same time, AI Agents and AI-Ready Data are climbing the peak of inflated expectations.

There’s huge excitement and also a clear difference in maturity:
👉 AI Agents will probably plateau sooner, evolving alongside GenAI as we learn to make them more reliable.
👉 AI-Ready Data, however, is a long game. Making biomedical data clean, connected, and harmonized, especially as new data types keep emerging, will take time and collaboration across the ecosystem.
And then there’s a quiet success story: Knowledge Graphs.
They’ve moved past the hype, sitting on the slope of enlightenment where technology actually works and delivers value.

Today, knowledge graphs are helping us connect complex biomedical knowledge and make AI systems smarter, context-aware, and traceable.

Right next to them, Model Distillation is maturing too, enabling smaller, specialized models that perform reliably in specific scientific domains.
It’s a nice reminder:
✨ The real breakthroughs often happen after the hype fades.
✨ Sustainable progress comes from building foundations including data, structure, and trustworthy models.

In the coming weeks, I’ll be sharing more about how knowledge graphs and model distillation are shaping the next phase of intelligent systems in life sciences.

What about you? Which of these AI trends do you think will make the biggest impact in your field?

#AI #DrugDiscovery #Pharmacovigilance #KnowledgeGraphs #ModelDistillation #GenerativeAI #Gartner #AIInnovation