Rules‑based AI and natural language processing
Rules‑based AI and natural language processing:
extracted structured insights from unstructured data (e.g., clinical notes)
learn more
Machine learning models
learn more
Machine learning models:
enabled more predictive and pattern‑based analysis, expanding scale and speed
Generative AI (GenAI)
learn more
Generative AI (GenAI):
introduced interpretation, creativity, and variability—along with risks like hallucinations
Agentic AI
learn more
Agentic AI:
represents the next disruptive phase, where AI systems break problems into steps, reason through tasks, and even signal uncertainty
Darmendra Ramcharran, senior director of Safety Innovation & Analytics for GSK, says AI is not a single technology and its rapid evolution is the source of its disruption. Mr. Ramcharran breaks down AI’s evolution into four phases, where importantly, each phase adds options, rather than replacing technologies of earlier phases, further expanding the AI toolbox:
Breakdown