It is a good question. Drug development will benefit a lot but not equally at different stages. AI is massively useful for anything that is related to data-rich operations, especially with massive amounts of text/media. So, anything related to paperwork and trial-related regulatory processes — will be completely disrupted at time passes. Also, AI is great for biology modeling and so eary stage drug discovery will benefit massively too, with time. Speaking about manufacturing, though, I think the AI’s role will still be mostly functional. Predicting bottlenecks, systems vulnerabilities, process optimizations etc, but I don’t think AI can completely change the manufacturing as we know it. The close we are moving to physical operations (moving atoms part) the less AI will be of an impact I guess. So in my opinion, the biggest role of AI will be in clinical trials design, diangnostics, administrative aspects, regulatory filings. Also, early stage biology research (target discovery, disease models, etc). Experimental and manufacturing part will mostly remain the same.