Anthropic and Citrini test AI on scorpion toxin binders
A collaboration between Anthropic and Citrini Research used Claude to generate protein candidates for scorpion neurotoxins. The experiment cost under $200,000 but highlighted the gap between digital speed and biological verification.
Financial research firm Citrini Research collaborated with Anthropic to test AI capabilities in drug discovery, specifically targeting scorpion neurotoxins. The team used a modified 35-page prompt to direct Claude, deploying 300 subagents that generated 318,000 protein candidates. After independent laboratory testing, 17 of these candidates were found to bind to four different scorpion toxins, although none have yet been shown to neutralize the venom. The entire process cost under $200,000, a fraction of traditional research costs. Citrini initiated the project to explore asymmetric opportunities in life sciences, a sector it viewed as undervalued.
The experiment highlighted the gap between digital speed and physical biological constraints. While AI compressed the design work into days, the full timeline to release remains 4 to 7 years due to synthesis and testing requirements. Anthropic supported the effort because scorpion stings affect 1.2 million people annually, causing approximately 3,200 deaths, a market too small for traditional pharma incentives but addressable by AI-reduced costs. The collaboration demonstrates that AI can accelerate early-stage biological research for neglected diseases, but physical verification remains a significant bottleneck that prevents immediate clinical impact. As one participant noted, virtual biologists may have compressed a year's worth of work into a couple of days, but biology works at the speed of life.