Today was a loud one. Meta said it is cutting roughly 8,000 jobs while raising AI spending to $145 billion. Google and Blackstone announced a joint venture that could plow up to $25 billion into AI data centers. And Novo Nordisk handed OpenAI the keys to a chunk of its drug pipeline.
If you squint, it is all the same story. The money is moving into machines and away from people. Worth thinking about who that math actually works for.
Meta Trades 8,000 People for Bigger AI Bills
Meta is laying off about 10 percent of its workforce and scrapping plans to fill another 6,000 open roles. Same week, the company bumped its 2026 capex guidance by up to $10 billion, landing at $145 billion for the year.
Why it matters: this is the cleanest version yet of the trade Big Tech keeps making. Take the salary line down, take the GPU line up. The cuts pay for the chips.
My take: I do not buy the line that this is an efficiency drive. It is a bet. Meta is betting that the AI it ships next year is worth more than the 8,000 humans it just sent home. If that bet is wrong, those folks are not coming back and the stock takes the hit anyway. Brutal trade either way.
Google and Blackstone Drop $25 Billion on AI Boxes
Google and Blackstone set up a joint venture to build AI infrastructure. Blackstone is putting in $5 billion of equity to start, with a 500 megawatt target by 2027. With leverage the total could hit $25 billion.
Why it matters: a private equity giant just got into the data center landlord business at scale. The hyperscalers do not have enough power or enough buildings on their own, so Wall Street is filling the gap and clipping a yield on it.
My take: this is the part of the AI boom that does not feel like software. Concrete, transformers, substations, and a 30 year lease. The folks who win this round own dirt and amps, not models. That is a real shift in who profits.
Novo Nordisk Hands OpenAI the Drug Pipeline
Novo Nordisk, the Ozempic company, signed a strategic partnership with OpenAI to push AI into drug discovery, clinical trials, manufacturing, and supply chain. The pitch is faster paths to new obesity and diabetes meds.
Why it matters: pharma has been the quiet AI use case nobody could prove out. Putting a brand the size of Novo Nordisk behind a wall to wall OpenAI deal is a signal to every other drugmaker that sitting on the sidelines is the risky move now.
My take: I want this one to work. Faster, cheaper drug discovery is the kind of AI promise that pays off for normal people instead of just shareholders. But OpenAI inside a pharma stack means the model is touching trial data and patient information. That deserves more sunlight than a press release.
The Through Line
Three different stories, one balance sheet move. Cut headcount, buy concrete, rent a model. We are watching capital reorganize itself in real time around AI, and it is not pretending to be subtle about it anymore.
Keep an eye on the next earnings cycle. The companies cutting hardest are also spending hardest. One of those lines has to start producing revenue soon or this whole thing gets uncomfortable.