Why Digital Twins Won't Replace Patients | Nindhana Paranthaman & Paul Agapow

AI arrived in drug development carrying a very large promise. Nindhana Paranthaman, Vice President of Early Clinical Development at Bicara Therapeutics, and Paul Agapow, Chief Scientific Officer at Bayezian, work on opposite sides of that promise and agree on more than the marketing would suggest. Paranthaman names the gains that actually arrived, all of them operational, and the one that did not: more drugs advancing into the clinic. Agapow explains why a trial completed inside a year is marketing, takes the digital twin apart on logical rather than technical grounds, and estimates that about a third of what the industry is building today will still be standing in five years. They close on target herding and what an investor should ask an AI-first company. (01:24) What AI actually delivered in clinical development(05:24) Whether a one-year trial is marketing spin(10:30) Digital twins, and why a twin cannot replace a patient(16:39) What survives five years, and the breakdown(18:48) Protocols, amendments and the unglamorous gold(29:29) Why biology breaks AI: discovered, not invented(36:35) Target herding and the hundredth me-too drug(39:09) Advice for VCs, family offices and private equity