Hackquisitions & Hackquihires

The deal is done. Meta is paying $14.3B for a 49% stake in Scale AI. Of course, that's not really what they're buying. They can't exactly come out and say it, but it's all right there in the reporting and the various releases and statements from the companies and individuals involved. What Meta is really paying for in the clear order of importance to them: Scale CEO Alexandr Wang, some of Wang's top employees at Scale, the stake in Scale itself.

Actually, there's one thing above those three that Mark Zuckerberg clearly believes he is buying here: relevance in the future of AI.

On the surface, this deal is a wild one for a wide range of reasons, even beyond the price tag. But to anyone who has been following the tech M&A landscape over the past year, this type of deal should be oddly familiar. It's a "hackquisition" – an acquisition of a company in all but name. Because there's no way the US government (and undoubtedly many other governments around the world) would allow such a deal to take place – yes, even in our post-Lina Khan world. So it's not an acquisition, you see. It's an agreement between two consenting companies to work together in weird and decidedly expensive ways while perhaps swapping an employee or two – or basically all of them, in some cases. Notably, always including the CEO. Nothing to see here. Nothing at all...

This latest version of that type of deal, pioneered by Microsoft with their deal totally-not-a-deal to buy totally-not-buy Inflection (an idea that itself was likely born out of the deal they almost had to do to bring Sam Altman and the OpenAI team on board post-coup), is the most interesting yet. And not just because it's the largest one yet. But because it seemingly blends together elements of past hackqusitions. In a way, many of the past such deals were really "hackquihires" and this is much more of an actual "hackquisition".

Let's look at the previous deals...