I don't want to praise Zuckerberg, but Muse is indeed a smart move

In Silicon Valley, whenever Zuckerberg prepares to deliver a passionate sermon on the next generation of technology, Wall Street's first reaction is usually to flip through Meta's capital expenditure sheet.

From the metaverse to AI, the script has barely changed.

Especially heading into 2026, Meta continues to purchase GPUs at scale, build data centers, and poach model and Agent talent on one hand, while pushing its full-year capex guidance to $130–145 billion on the other. Once spending reaches that magnitude, the secondary market naturally wonders, how exactly is all this money going to be earned back?

The old standard answer was simple enough — feed the models into the recommendation algorithms, make ad targeting more precise, and keep squeezing more efficiency out of the massive money-printing machines that are Facebook and Instagram.

The problem is, that logic has a ceiling too.

For a giant with hundreds of billions of dollars in annual revenue, if over a hundred billion in AI investment ultimately just makes feed ads slightly more accurate and click-through rates slightly higher, why should the market grant it new valuation imagination?

That's what makes Muse interesting.

For the first time, it connects Meta's increasingly massive AI investment over the past few years to a product that sits outside advertising yet remains close enough to real consumption and transactions. From model capabilities, to Agent products, to transactions and monetization, at least in logic, a complete path is beginning to emerge.

This is also why I really don't want to praise Zuckerberg, but I have to say, this move with Muse is extremely shrewd.

1. Meta's AI Is Not Just a Cost Center

Let's first look at the recent market action.

From late August to September 21, Meta's stock rose from over $570 to $741.25, a cumulative gain of nearly 30%, with September 21 alone surging 11.34%.

Of course, attributing all 30% to Muse would be inaccurate. Meta's rally had already started before Muse's launch, and the earlier move was more about repairing the market's pessimistic expectations around capex, cost growth, and margins.

But after Muse went live on September 8, discussions around Meta AI did gain an extra dimension: could this money actually grow a second growth curve?

That's where the difference lies.

Muse was not built on the chatbot logic at all. It is a very pure personal AI Agent — once you grant it authorization, it can jump into web pages on its own, call APIs, fill out forms, compare prices across platforms, and even when you lock your phone screen, it keeps running errands for you in the background, only pulling you back in to confirm at the moment of payment or sending a critical email.

In other words, AI in the past was more like answering "what should I do," while what Muse wants to do is take one step further — "okay, I'll go do it."

This shift may look like just moving from Chatbot to Agent, but the commercial significance is entirely different. After all, once AI can take over actions, it naturally cuts into search, shopping, bookings, and payments — the entry points closest to money.

And this is precisely where Meta truly deserves a re-rating.

It doesn't necessarily need to make Muse the world's number-one benchmark large model. For the vast majority of ordinary users, they don't care whether the model scores two points higher or three points lower — they only care whether it can actually get something done after they hand it a task.

This echoes our judgment on Google from a few days ago. As the capability gap between underlying models begins to narrow, companies that truly own entry points, product ecosystems, and distribution capabilities are actually better positioned to capture the dividends of AI's second half (further reading: "Gemini 4.0 Is So Strong It's Flooding Feeds — Is Google's AI Counterattack Here?").

From this angle, Muse is not Meta re-entering an exam called "whose large model is strongest." It simply switched exam rooms: when underlying model capabilities are basically sufficient, I'll be the first to quietly slip an Agent into the daily lives of my billions of users.

And in this exam room, the hand Meta holds starts to look very different.

2. Intent and Interception: Why Is Amazon Guarded While Shopify Embraces?

To do this, the cards Meta holds are something others can only envy — native distribution.

AI startups fret daily over customer acquisition costs, while Meta sits on Facebook, Instagram, WhatsApp, Messenger, plus its still-advancing smart glasses. Billions of people are already immersed in its products.

But if it were just "lots of people," that alone wouldn't explain Muse's value.

What's truly interesting is that it has the opportunity to push what Meta already controls from "Attention" further down to "Intent."

For example, previously when we scrolled past a soy milk maker video on Instagram and paused for a few seconds, the algorithm might judge that you're recently interested in kitchen appliances, so it shows you a few more related ads, and Meta earns the impression fee.

But if you actually want to buy, you'll most likely close Instagram, open Amazon/Temu, search, compare prices, read reviews, and place an order.

In other words, Meta grasps "what you might like," but the actual consumption decision and transaction still happen on someone else's turf.

Muse, however, could completely restructure that path — we simply tell Muse: "Find a soy milk maker under $100, easy to clean, suitable for a couple," and the subsequent filtering, comparison, confirmation, and payment all run through the Agent's channel.

Between attention and purchase intent, it looks like just one step, but in the business world, that step is worth a lot of money.

Understanding this explains why major tech companies have such polarized attitudes toward Muse.

Amazon is on high alert, quickly restricting Muse from placing orders on customers' behalf, citing authorization compliance and privacy security, while Shopify is rolling out the green carpet, eager for Shop Pay to be embedded directly into Muse's transaction flow.

It's purely determined by where their interests sit:

• Amazon itself is a closed center of search and transactions. Its most valuable asset isn't cloud services, but the moment when consumers must open Amazon to make a purchase decision. If everyone in the future buys things through a single instruction to an external Agent, Amazon would be forced into becoming a pure back-end shelf, and its painstakingly built on-site search bidding and advertising system would have its ladder pulled out entirely;
• Shopify, by contrast, has always been a water seller. Whether orders come from Instagram, from Google search, or drop in from the Muse Agent, as long as they ultimately settle through its payment system and go through its merchant channels, it can steadily take a cut;

This is the most realistic power restructuring of the Agent era. Whoever is closest to the user's final decision controls the "Strait of Hormuz" on the e-commerce map.

Of course, we shouldn't be blindly optimistic here. Many people scroll IG, but those willing to entrust their email passwords, calendar permissions, and credit card numbers/CVVs to an AI are still a minority at this stage.

After all, if GPT or Gemini gets something wrong once, it's just AI having another hiccup, but if an Agent buys the wrong quantity when placing an order for you, books the wrong hotel, or sends an email that shouldn't have been sent, the user's trust in it could be gone in one shot.

So if Meta wants to eat the fattest piece of meat that is "intent," it first has to solve a very basic problem: whether people dare to actually hand things over to it.

3. Subscriptions Are Just Small Change — the Real Big Business Is in Transactions

Ultimately, we still have to do the math. How does Muse help Zuckerberg recoup?

The easiest thing to understand right now is subscription fees. Muse offers a free version, as well as paid tiers at $20 and $100 per month. Assuming it could eventually get 10 million users paying $20 per month long-term, that's $2.4 billion in annual revenue.

For most software companies, that would already be a very large business, but for Meta, it doesn't even begin to quench the thirst.

After all, Meta's Q2 revenue alone this year hit $60.8 billion, with advertising accounting for $59.3 billion. Selling memberships alone simply cannot fill that massive hardware black hole.

So in the view of Maitong Research Institute, subscriptions are not the most noteworthy part of Muse. What truly has imagination is transaction commissions — that's the main course.

If an Agent becomes the digital butler for hundreds of millions of people, the gross merchandise value (GMV) it撬动 would theoretically far exceed its own membership fees. At that point, the monetization space opens up completely: transaction matching commissions, traffic fees from merchants wanting to rank higher in recommendations, revenue sharing on payment channels...

In the past, merchants advertising on Meta were paying for an exposure probability that "you might buy." But the Agent of the future holds directly the certainty of orders where "the user is already ready to pay." So would merchants be willing to pay commissions for completed transactions? Would payment companies be willing to share revenue for Agent Checkout? Would merchants be willing to purchase some kind of commercial service from it to become a priority option?

These are the things more worth imagining.

However, this also plants a landmine for Meta.

The reason people use your Agent is because they believe it's on their side, a "silly girl assistant" maximizing their interests. But if one day Muse frantically recommends a certain hotel, and people discover it's not because that hotel is suitable, but because it offers Meta the highest rebate, then the trust foundation of this product collapses on the spot.

So for Meta, the deeper monetization goes into the transaction end, the more it must handle a contradiction: how to make money from user intent without making users feel their intent has been sold.

This is also why Meta must repeatedly emphasize that Muse's private conversations and Secure VM will not be directly connected to the advertising system — it must stand on its own as a new charging and credit system outside of advertising.

Why this matters so much ultimately comes back to Meta's increasingly exaggerated AI bill. Revenue grew 28%, but total costs and expenses surged 55%, operating margin was slammed from 43% down to 31%, quarterly operating cash flow was $31.8 billion, while capex plus lease principal alone consumed over $31 billion, squeezing free cash flow into an extremely narrow range of $784 million.

Although one-time charges are mixed in, one fact is already very clear: Meta is reinvesting cash back into AI infrastructure at an unprecedented pace.

If all these GPUs burn through to the end and merely add a few more clicks to feed ads, then at best it's patching the existing business and cannot support the next round of valuation restructuring.

But if Muse truly tears open a transaction entry point led by Agents, then Meta's story in the eyes of the capital markets could leverage AI to build a new entry point, completely transforming from "an ad intermediary frantically buying shovels" into "the dispatch console for next-generation internet life."

These are two completely different valuation models.

Final Thoughts

It must be acknowledged that for Meta, Muse is beginning to connect those increasingly frightening AI bills of the past few years to a sufficiently concrete commercial path.

What Meta possessed in the past was the world's largest attention machine; what Muse wants to do is follow attention further down — from knowing what you like, to understanding what you truly want, to getting things done on your behalf.

If this chain truly runs through, what Meta gains may not just be a ChatGPT competitor, but a new layer of internet entry point.

Of course, the stock price has already risen to over $700, and the market has already offered considerable applause in advance.

Wins and losses on model leaderboards may soon pass, but how this account gets settled is what's truly worth trading in META's next phase.

Followofficial X for US stock news

Joinofficial community to discuss trending topics