Pharma Sessions

The episode contains practical advice on AI adoption and examples from real work (COVID demographic analysis, data science projects), but much of the content recycles familiar frameworks (automation of repetitive tasks, creative work as AI-resistant, people-centric approaches). The insights about data integration challenges and change management friction are moderately useful, but the conversation includes significant padding with anecdotes and general principles that experienced operators would already understand.

“focus on what your role is and look at it and say what things here are really repetitive that could easily be done by a machine”

“what I found with like enterprise applications is the research and development piece where you really are building unique applications for unique problems that are pharma specific. And these are the ones that are more challenging”

The core arguments - upskill in AI tools, focus on what machines cannot do, maintain human expertise - are standard across the AI-adoption discourse. The specific COVID demographic analysis example is concrete but represents a straightforward data-joining exercise rather than novel thinking. The book title itself (profit and protect) uses the familiar risk/opportunity binary. There is minimal counterintuitive or first-principles thinking presented.

“focus on what your role is and and look at it and say what things here are really repetitive that I that could easily be done by a machine”

“are you on the right track? Yep, is hyper valuable. And that's where I'm still very um people-centric”

Tim Smith is credible: head of data science at Takeda, early ML adopter in pharma (2008 Novartis work), published author on AI. He has built commercial applications and led institutional collaborations. However, the transcript reveals limited depth on his most recent strategic work at Takeda - most substantive examples are from earlier career stages (Novartis, startups). His role emphasizes change management and adoption rather than cutting-edge technical or discovery outcomes, which reduces practical relevance for R&D operators.

“head of data science community and AI innovation at Takeda”

“I started at Novartis. I pretty quickly got, they realized I like to do sort of large-scale projects. So building like pipeline analytics, a database to track all of our projects across the company”

The episode offers some concrete examples (COVID demographic analysis with CIA data, Sunshine Act data mining, flow cytometry lab sharing tool, Meta layoffs mentioned in news) but lacks quantitative depth. Most claims are illustrated through anecdote rather than data: no metrics on adoption rates, ROI timelines, cost savings, or clinical/commercial impact. The discussion of data integration challenges is conceptual rather than case-studded; the legal profession examples are generic.

“countries, Lesotho, which is like 50% under the age of 18, they almost had no infections. And then older countries had many more”

“the Sunshine Act data, which companies are paying which researchers and and kind of building trends and stuff like that from that data”

The host asks reasonably structured questions and attempts follow-ups (e.g., probing change management bottlenecks, data integration challenges), showing genuine curiosity. However, the conversation rarely pushes back or tests claims. When Tim makes broad assertions (e.g., 'people are still the key,' medicinal chemists remain essential), the host largely affirms rather than challenge or ask for evidence. The karaoke icebreaker and personal anecdotes consume air time without advancing substance. The host does eventually sharpen focus on data normalization challenges, but overall the dynamic is more collaborative than critical.

“And a big part of your role is around institutional change and enter enterprise change and adoption of this. So, what are you seeing around change management?”

“Well, and then there's also even once they understand it, what I've seen, and I don't know if this is improving or not, but there are real sometimes operational challenges”