Big Pharma Can Rival Big Software
Summary
- Eli Lilly's market capitalization exceeds $1 trillion, driven by its GLP-1 portfolio which generated nearly $17B in Q3 2026 & is projected to reach over $100B in annual sales by 2030.
- The pharmaceutical industry operates on a dual economic model: a "hits business" where the top 25% of drugs generate ~85% of revenues, & an "annuity business" where successful drugs provide long-term cash flow under patent protection.
- Over 300 drugs are projected to be "blockbusters" generating >$1B annually, with the top 10 pharma companies deriving ~50% of revenue from just their top 3 assets.
- Traditional drug development is costly & slow, with a present value cost of ~$2B & an average timeline of 13 years, while less than 10% of candidates entering human trials gain approval.
- AI aims to transform pharma economics by increasing the probability of success & decreasing development time, potentially reducing the risk-adjusted cost of bringing a drug to market.
- The current market is approaching a GPT-3 level of maturity for bio-AI, characterized by improved scientific reasoning via LLMs & advanced molecule generation using diffusion-based models.
- Major pharmaceutical companies are investing heavily in AI partnerships, with deals involving Isomorphic Labs, Chai Discovery, Boltz, & BigHat Bio totaling billions in potential milestone payments.
- AI-native business models are emerging, ranging from selling data, software, & predictive models to pharma giants to building full-stack biotechs that discover & develop their own drugs.
- Significant challenges remain in applying AI to biology, including data scarcity, system complexity, & long, non-verifiable feedback loops requiring wet lab validation.
- The industry is shifting toward valuing the strength of a company's discovery engine & AI infrastructure, moving beyond traditional metrics that focus primarily on approved drug portfolios.
AI Will Make Pharma Even Bigger
- Keywords
- AI in Bio + Health, Bio + Healthcare, Growth