Can America Win The AI Biotech Race Against China? | Lada Nuzhna & Elliot Hershberg

Episode Summary

Executive Summary: The conversation argues biotech is in a paradox: science and AI are advancing rapidly, but economics, regulation, and trial infrastructure have made drug development slower, costlier, and less investable. The guests see China as a speed/cost competitor, AI as a major but not yet sufficient fix, and new modalities plus aging/longevity as the next big opportunity for iconic biotech companies.

Main Topics: Biotech’s economic downturn vs. scientific progress (Priority: 5/5): The hosts contrast weak public-market performance, few IPOs, and EV-negative companies with major scientific progress such as AI-designed antibodies and advances in genomics, concluding the industry’s business model is under strain even as the science improves. Regulation, trial costs, and industry inertia (Priority: 5/5): The guests argue that the cost and complexity of clinical development have risen dramatically due to regulation, CRO consolidation, and entrenched processes. They suggest the bottleneck is not only FDA rules but also executional and cultural inertia across the trial ecosystem. China as a competitor on speed and cost (Priority: 5/5): China is presented as a growing biotech challenger that benefits from faster trial approval, lower costs, and a more efficient investigator-initiated trial system. This threatens the traditional U.S. model where America invents and the rest of the world scales. AI as a biotech multiplier, not a full solution (Priority: 5/5): AI is expected to become ubiquitous in biotech within five years, but the key question is whether it can materially reduce cost and timelines, improve human efficacy prediction, and generate novel molecules and platform products rather than only optimize preclinical work. New modalities and platform companies (Priority: 4/5): Both guests argue the next generation of biotech winners will come from new modalities and modern infrastructure, not just incremental target hunting. They point to generative biology, synthetic biology, genomics, and delivery platforms as sources of outsized value creation. Aging, longevity, and GLP-1s as catalyst (Priority: 4/5): Aging is framed as a massive but under-incentivized indication because payers do not reward prevention well. GLP-1s are cited as evidence that big, injectable chronic therapies can succeed and may reopen ambition for age-related disease and longevity drugs. Reforming incentives and commercialization pathways (Priority: 4/5): The discussion ends with proposals for better regulatory incentives, lower per-patient trial costs, faster U.S. first-in-human studies, and new designation-like support for chronic diseases analogous to orphan drug incentives.

Key Arguments: Biotech economics worsened because regulation and clinical execution got more expensive over time, while discovery science improved. The true bottlenecks are a mix of regulation, CRO/industry structure, and cultural assumptions that trials must be slow and expensive. China is no longer just manufacturing copies; it now competes on novel biology, gene therapy, gene editing, and very fast clinical execution. AI will be broadly used across biotech, but its biggest value will come from better efficacy prediction, better molecule design, and enabling entirely new product classes. The next iconic biotech companies will likely be platform/modality companies that bundle biology, modeling, sequencing, and delivery into new product architectures. Aging is a huge unmet need, but reimbursement and approval systems do not yet reward prevention or lifespan extension directly. GLP-1s show that the market can support large chronic-disease products and may shift the industry back toward big indications rather than only rare diseases. A more innovation-friendly U.S. trial system would preserve domestic invention and commercialization rather than pushing startups abroad for first-in-human studies.

Data Points: Trial cost per patient at Regeneron start: about $10,000 - Used as a historical baseline for how cheap early clinical trials once were. Trial cost per patient today: about $500,000 - Cited as the ballooned cost of dosing a patient in a trial. Industry spending per approved drug: more than $2 billion - Used to illustrate escalating R&D costs and declining efficiency. Public biotech below cash balances: one-fifth of public biotech companies - Described as trading at or below cash, showing market distress. No biotech IPO window: 7 to 8 months - A stretch with no biotech IPOs, highlighting public-market freeze. Trial geography for Amplify companies: 0 of 3 planned first-in-human trials in the U.S. - Example of startups going abroad for initial studies. Chinese IND review model: 30 days - Implied approval if no proactive hold is issued after filing. China investigator-initiated trial review: 5x to 6x faster - Used to explain why many novel trials go to China. Orphan drug approvals in 2024: 50% - Cited to show how incentives can dramatically shape drug development. Approved orphan drugs before Orphan Drug Act: less than 40 - Historical comparison of pre-incentive drug development. Human lifespan benchmark in Japan: close to 80 years - Used as an example of lower heart-attack mortality and longer lifespan. Caloric restriction in monkeys: about 2.5 years added to a 25-year median lifespan - Discussed as the only well-documented lifespan effect in monkeys. High-throughput screen efficiency: about a billion times more efficient than 20 years ago - Illustrates dramatic scientific progress despite rising development costs.

Pivotal Quotes: "There is no law of physics that requires it to be $500,000 in terms of complexity and cost to dose a patient in a trial." — Elliot Hirschberg: On why clinical development has become unnecessarily expensive and could be improved. "We have to invent stuff." — Elliot Hirschberg: On how the U.S. biotech industry can compete against China by creating genuinely new modalities and biology. "Everyone will be using AI in the biotech industry five years from now." — Unnamed speaker in transcript: On AI becoming a standard tool across the biotech stack, though not a complete solution.

Implications: Biotech’s future likely belongs to teams that combine novel modalities, better data, and faster execution. Winners will reduce trial friction, build platform businesses, and target huge indications like aging and metabolic disease.