🧠 Food for thought

Implications, context, and why it matters.

Lilly grows AI in-house and through partners

• Eli Lilly is taking a two-track approach, it is building internal AI capabilities while also partnering for outside innovation 1.
• Recent moves include a joint AI lab with NVIDIA (a leading maker of AI chips). The companies said they will jointly invest up to $1 billion over five years, and Lilly has launched “LillyPod,” an AI supercomputer it says is powered by 1,016 NVIDIA Blackwell Ultra GPUs (graphics processing units, chips commonly used to train and run AI models) 12.
• Insilico brings a track record, it has developed at least 28 drugs using generative AI tools, with nearly half at a clinical stage 3.
• Other Insilico agreements include a deal with Servier (a French pharmaceutical company) that Pharmaphorum said could be worth up to $888 million plus a Menarini deal valued at over $550 million including milestones, which supports the case for its AI platform across the industry 45.

AI discovery shifts toward licensing drug candidates

• The partnership moved from a software license in 2023 to a deal for AI-discovered drug candidates, which suggests AI is becoming a source of licensable assets 3.
• This setup lets AI biotechs act as efficient R&D (research and development) engines that reduce risk for molecules before large drugmakers take over.
• Insilico has said it can reach preclinical candidate nomination in an average of 12 to 18 months. The company cited a 4.5-year traditional industry average benchmark 6. This can keep startups focused on discovery while later trials take more capital.
• The split keeps AI development outside China, while early preclinical work happens in China. That structure can limit geopolitical risk while still using global R&D capabilities 3.