Why Pharma’s AI Bet Might Be Wrong | Martin Shkreli

· Source: Weights & Biases · Field: Health & Wellbeing — Pharmaceuticals & Biotechnology, AI in Drug Discovery · Depth: Intermediate, quick

Summary

Martin Shkreli argues that artificial intelligence (AI) has a minimal role in drug discovery, contrary to widespread industry belief. He contends that the primary bottleneck in drug discovery is the initial idea or molecular target selection, citing Dupixent as an example where target identification was the critical step, not the subsequent antibody or small molecule development. Shkreli asserts that making antibodies or small molecules is a well-established process, representing only about 10% of the overall drug invention. The remaining 90% involves human testing, an area where AI offers little assistance. His analysis, presented during a livestream, identified only very small parts of the drug discovery system where AI could genuinely contribute.

Key takeaway

For investors evaluating pharmaceutical companies heavily invested in AI for drug discovery, you should critically assess where AI is being applied. Focus on whether the AI addresses the fundamental bottleneck of novel target identification or merely optimizes later-stage, well-understood processes like molecule synthesis, which Shkreli suggests offers limited value. Your due diligence should differentiate between genuine innovation and incremental improvements.

Key insights

AI's role in drug discovery is limited, with idea generation and human testing as major bottlenecks.

Principles

Topics

Best for: Investor, Entrepreneur, AI Product Manager, Executive, Domain Expert

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Weights & Biases.