On massive AI acquisitions, robotics, and headcount trimming

In the age of AI, are acquihires more valuable than acquisitions? In honesty, that question presents a false binary. Acquihires and acquisitions are both tools in the strategy toolkit. The key is knowing when to use each, and how to align them with the long term trajectory of AI. 

Some years ago, acquisitions often meant buying networks, distribution, and monetizable users (think Instagram, WhatsApp). But in AI, the most valuable assets are compute, data, and sometimes people. Many of these latest “acquihires” optimize for horsepower or elite strategists, not existing users. The playbook has shifted. 

Zuckerberg’s recent acquisition of the Scale AI team, including its CEO Alexandr Wang, is a great case study. After years of being on the defensive, Meta is waking up to the reality that open-sourcing LLaMA isn’t enough. 

The Scale AI stake—and the rumored hiring of Nat and Daniel—are huge moves to reboot Meta’s long-term narrative in AI, getting back into the race with OpenAI, Anthropic, and Google DeepMind. 

What Zuckerberg—and so many others—are buying is optionality. In the case of Scale AI, he’s folding in a team that knows how to train large models with the most efficient use of human feedback loops. In the case of Nat and Daniel, it’s a braintrust that knows how to architect and operate AI-native companies.

The winners in the AI race aren’t necessarily the ones who build smarter tools. In my opinion, the true breakthroughs will come by reorganizing entire companies, economic sectors, and even geopolitics around cognitive infrastructure. Acquihires are just one strategic move in the game of those who acquire capital, compute, and talent faster than anyone else.

Spotify, Apple, et al.: https://link.chtbl.com/AGKXHA-Y

YouTube: https://youtu.be/hWU7x9eVyD0

Transcript: https://www.possible.fm/podcasts/riffs026/

You can subscribe to catch more episodes of Possible here: https://www.possible.fm/podcast/