This is just the beginning…
Microsoft is offering voluntary buyouts to about 60,000 people, roughly a quarter of its workforce, and Skydance is running buyouts of its own on the same day. These are not the same story, and that is the whole point.
Multiple things are happening at once, and multiple things can be true at once.
Microsoft first. CNBC reported it today. The company is offering voluntary exits to around 60,000 employees across the US, Canada, the UK and India, as it tries to balance a costly AI infrastructure build out against capital efficiency. Microsoft had about 223,000 full time employees as of June 30. The offer runs up to level 70, which covers general managers and some vice presidents, and the deadline is the first week of December. Worth noting that some groups including sales are not participating, and Microsoft says this does not replace layoffs.
Then Skydance. Same day. Voluntary buyouts plus a five day in office mandate, announced by David Ellison and co-CEO Ynon Kreiz. That extends a Paramount policy to the Warner Bros. Discovery operations they just absorbed, where staff had been on a three day hybrid schedule.
On the office piece, during COVID a lot of people assumed remote was a clean win because you save on office space. What the studies have shown since is that depending on the industry it often did not help in the long run. You are going to see more of this.
But the Skydance and EA situations are not the Microsoft situation.
When you have a deal that size, and a party that has to take on a mountain of debt to make it happen, new leadership comes in, looks at the bottom line, and starts cutting. Paramount closed the Warner Bros. Discovery acquisition on October 6 for around $110 billion, and Ellison and Kreiz have identified roughly $6 billion in cost savings they intend to find. That is what these buyouts are. Same with EA after the Saudi led buyout. That is debt math, not an AI story.
Microsoft is the one where AI is plausibly a real factor, because they have poured an enormous amount of money into it. If the economics shift enough that work is more profitable done a different way, that shows up in headcount eventually.
There is also the 2020 era correction nobody wants to name. A lot of industries ate off that boom and overhired badly. Some of what gets called AI efficiency is a company finally admitting it overstaffed four years ago. I do not think that is the whole story. I also do not think it is nothing.
Where we actually find out what AI is worth is when the investment starts catching up. Right now this is a speculation market. The biggest players are not getting a return, and the pressure to show a profit grows every year. When investors come to collect, we will see what corrects.
Google and Microsoft have the advantage of making money other ways, which cushions what they spend on infrastructure. The companies that are all in on this alone have balance sheets that look nothing like that.
Something being true today does not make it true tomorrow. That is the white pill in it. I do not know how this plays out, and neither does anybody else.