Background: Meta is the parent company behind Facebook, Instagram, and WhatsApp, along with some of its less successful bets, including Oculus headsets and the short-lived Metaverse. But over the past couple of years, Meta has spent tens of billions on AI infrastructure, without it being clear where that spending was actually going.

What happened: Meta launched its AI personal assistant, Muse, in the US, two weeks ago. It rocketed straight to the top spot on the US App Store... and investors liked what they saw. Meta's share price jumped 11%. In fact, some Wall Street investors reckon Muse could add nearly US$30 billion in revenue for Meta by 2030.  

What else: Muse is less chatbot, more AI agent. It sends emails, books travel, shops online, and fills out forms. And while Meta does have paid subscription options, that's not where the really juicy opportunity lies.

What's the key learning?

💡 Giving something away for free is usually the first move in a much bigger game. Right now, Meta is giving away free tokens and free compute, betting that the value comes from what users do inside the ecosystem.

💡 Meta's framing is that Muse should "make you money and save you money." With 10 million+ active advertisers already on Facebook, Meta doesn't need to charge $20 a month if Muse can drive more hundreds or thousands of dollars in more ad spend. And if it starts recommending and buying products for you as a consumer, Meta could take a cut there too.

💡 Google uses the same playbook with Search and Gmail, giving them away for free while building an ecosystem that generated US$264 billion in advertising revenue last financial year. And investors clearly seem much more excited about this AI opportunity than they were about the Metaverse.