Novo Nordisk Hands the Keys to OpenAI
Novo Nordisk announced a top to bottom partnership with OpenAI. Drug discovery, clinical trials, manufacturing, supply chain, commercial ops. The whole shop. They want it rolled out across the business by the end of the year.
Why it matters. This is not a pilot. This is not a sandbox. A company that makes the most talked about drug of the last five years is bolting AI to every floor of the building. When pharma moves, the rest of the regulated world watches.
My take. Novo is losing ground to Eli Lilly and they know it. Speed wins right now. If AI shaves six months off a trial design, or finds one promising molecule the lab missed, that pays for the whole deal twice over. The smart play is being made out of fear, not vision, and that is usually when companies actually move fast.
OpenAI Wants One Super App to Rule Them All
OpenAI is folding ChatGPT, Codex, and the developer API into one product team. The plan is a single super app, with the Atlas browser pulled in for good measure.
Why it matters. Up to now OpenAI was a research lab that happened to sell three or four products. This is them finally acting like a product company. One front door, one bill, one experience.
My take. I get it. Splitting your best engineers across four overlapping surfaces is how you lose to a focused competitor. But "super app" is a phrase that should make you nervous. Tencent pulled it off in China. Facebook tried it and got Meta. The graveyard is full of products that wanted to be everything for everybody. OpenAI has the muscle to try, but they better not let the chat experience get worse so the API folks can ship a feature.
The Air Force Wargames at 10,000x Speed
The Air Force ran its first real wargame using WarMatrix, an AI wargaming environment. More than 150 people, six 24-hour moves, physics based modeling, AI doing the adjudication. The system can run simulations up to ten thousand times faster than real time.
Why it matters. Wargaming used to be a room full of colonels arguing over a paper map for a week. Now it is a server running a million versions of next Tuesday before lunch.
My take. This is the quiet story but probably the biggest one. The first country that figures out how to plan and rehearse at machine speed will have a real edge over the country still doing it by hand. The optimist in me says better planning means fewer mistakes. The realist in me says faster wargames means faster wars. Both can be true at the same time.
What I Am Watching
Three stories, one pattern. Big institutions are done dipping their toes. Pharma, big tech, the military. They are all going whole hog this year.
If you are still treating AI like a side project at your shop, you are about to be running last year's playbook against people running this year's tools. Get on the field.