SAN FRANCISCO — Artificial intelligence could help American drugmakers compete more effectively with China, which has been bringing new drugs to market faster and at lower costs, Katherine and Erin write from the floor of JPM.

• Eventually.

State of play: No commercialized drugs exist that were fully developed by AI, or even ones whose development was AI-assisted.

• FDA approval of the first AI-enabled drug is likely two or three years away, HSBC Innovation managing director Jon Norris tells Axios.
• VC-backed Generate:Biomedicines might get there first: It is launching two Phase 3 trials for a severe asthma treatment engineered with AI.

Zoom in: McKinsey & Co. partner Bart Van de Vyver says most big pharma companies are prioritizing building AI capabilities in-house or partnering with AI platform companies, rather than pursuing outright acquisitions.

• Those platforms often generate stronger insights by aggregating data across multiple pharma partners.
• "Everybody benefits from the common algorithms," Van de Vyer says.
• So far, companies are mostly using AI to find biomarkers and identify disease and population targets, Maha Radhakrishnan of Sofinnova Investments adds.

Follow the money: AI hype is drawing generalist investors back to biotech, which could accelerate the revolution.

• It's also spawning a cottage industry of personalized care advocates, such as the one just acquired by OpenAI, which could be an AI application that patients use long before AI-designed drugs.

The bottom line: The Holy Grail would be medicine not only developed with AI, but also one where AI can successfully simulate human trials.

• JPM attendees don't expect that anytime soon, but do believe the industry is actively working toward it — beginning with target identification.