Google has more AI talent, more user data, and more infrastructure than Meta — yet Meta shipped the personal agent that Google should have built years ago.
The Summary
- Meta launched Muse, a personal assistant agent that handles purchases and bookings, prompting the obvious question: where is Google's version?
- The number two response to Muse (after privacy concerns) is "Why didn't Google build this?" — a question that reveals Google's decades-long focus problem
- Google is "undoubtedly working on it" but has multiple scattered AI agent products instead of one coherent offering
- Meta's advantage isn't technical capability — it's Zuckerberg's willingness to pick one direction and commit completely
The Signal
Google's failure to ship a Muse competitor isn't about capability. They have Gemini, they have Search, they have Assistant, they have Bard's successor products. They have the world's best AI researchers and more consumer data than anyone except maybe Amazon. What they lack, as MG Siegler points out, is focus. This is the same company that has launched and abandoned more messaging products than most companies have total products.
Meta's Muse is drawing two primary reactions: "I'll never trust Meta with my data" and "Why didn't Google build this?" The second question is more revealing than the first. It exposes the gap between Google's resources and Google's execution. They should have been first. They had every structural advantage.
"Say what you want about Meta as a company and Mark Zuckerberg as a leader, but they (and he) have focus."
Siegler draws a parallel to Meta's metaverse bet — the company went all-in, then pivoted hard when it didn't work. That's not indecision. That's how focus actually operates. You commit fully, you learn fast, you move. Google spreads bets across a dozen products that all do 70% of what users need and none of which become the default.
Stratechery notes that Meta also needs Walmart partnership to wait out Amazon's dominance in commerce, while Expedia is fighting to maintain its middleware position. These are real strategic challenges for Muse's ecosystem. But Meta shipped the product. They're learning in production. Google is still running internal pilots of seven different agent architectures.
Key structural differences:
- Meta has one AI agent product. Google has Gemini, Assistant, and scattered experimental features.
- Meta picked a CEO who can kill projects and reallocate. Google's structure makes that nearly impossible.
- Meta's revenue comes from ads on content they don't create. Google's comes from search traffic they increasingly disrupt with every agent feature.
The deeper issue is incentive alignment. Every time Google makes an agent better at completing tasks, they risk reducing search queries. Every completed booking through an agent is a click that never hits a search result page. Meta doesn't have this problem. Their revenue model isn't threatened by agents that work. If anything, keeping users inside Meta properties longer helps them.
The Implication
Watch for Google to either ship a unified agent product in the next six months or fracture further into feature sprawl. If they can't consolidate their AI agent strategy now, while Meta is still building Muse's partner network, they may never catch up. The window for "fast follower" is closing.
For builders in the agent space: the real competition isn't about who has the best models. It's about who can ship a focused product and iterate in public. Meta is proving that strategic clarity beats technical perfection. If you're building agent infrastructure or applications, focus matters more than your roadmap's completeness.