After experiencing Muse, I cleared out my Airbnb holdings

In Silicon Valley and Wall Street, discussions about how AI will reshape the business world have never ceased. But when theory turns into reality, the first to feel the chill may be those internet giant platforms we know best.

Recently, on the well-known financial podcast The Synopsis, host Drew had an in-depth conversation with independent veteran equity analyst Mostly Borrowed Ideas (hereinafter referred to as MBI), who has over 150,000 followers.

MBI shared that about ten days after Meta released Muse, he downloaded the app and began testing it. This test ultimately led him to make an investment decision—he liquidated all his heavily-weighted Airbnb stock and increased his position in Meta.

In this conversation, Drew and MBI also delved into how Meta’s AI Agent product Muse could disrupt the business models of internet aggregation platforms such as Airbnb, Booking, Uber, DoorDash, and Amazon. A brand new business era—“Proactive Commerce” may be about to unfold.

Left: Drew, Right: Mostly Borrowed Ideas

After Experiencing Muse, Sold Out of Airbnb

As a heavy Airbnb user (over $7,000 spent this year) and a major investor, MBI was initially dismissive of claims that AI would disrupt online travel platforms (OTAs), just like many others. After all, people need to see pictures, read reviews, and sometimes chat with hosts when booking accommodation; traditional text-based AI (like early ChatGPT) simply couldn’t meet these demands.

But Muse changed his mind.

MBI shared his experience: “I asked Muse to recommend five places to stay within a two-hour drive from my home, based on my past stays on Airbnb. Shockingly, Muse not only automatically browsed my history, but even analyzed travel short videos (Reels) I had saved on Instagram, precisely capturing my preferences. It worked like a real person, browsing and filtering in the background, and finally offered me several options, complete with rich images and reasoning.”

When MBI found a cottage he liked and asked if he could book directly, Muse told him within a minute: Yes, and you can book directly, bypassing Airbnb at 60% less cost. Muse had even linked a credit card, and with just one click from MBI, the payment would be completed.

“At that moment, I realized Airbnb may be on the wrong side of future trends.” MBI admitted.

Host Drew raised a challenge from the traditional business perspective: “Airbnb’s biggest moat is ‘trust’ and ‘exclusive supply’. As a consumer, I wouldn’t dare to trade directly with random strangers online—I need Airbnb as a middleman for refunds and after-sales, and many hosts don’t have their own independent sites.”

MBI countered: An AI Agent can easily build “trust” assessment by scraping genuine reviews from across the web. More importantly, in the age of AI, hosts only need to tell their AI once: “Help me sync this listing to Booking and VRBO, and manage the calendar.” The so-called “exclusive supply” barrier is instantly broken. Hosts have long suffered from single-platform “monopolies”—the friction cost of managing multiple platforms was high before, but AI eliminates this friction.

Airbnb vs Booking: Who Is More Resilient?

After sharing his Airbnb exit, the pair shifted focus to another online travel giant: Booking.

• The Curse of Direct Traffic: Airbnb currently has a whopping 90% “organic/direct traffic”—which is a key reason for its high profit margin. But in the age of AI agents, consumers might no longer open Airbnb’s app directly, but rather coordinate trips through AI. MBI predicts Airbnb’s direct traffic will sharply decrease in the next 5-10 years.

• Customer Acquisition Cost Shift: Currently, a third of Booking’s traffic comes from paid acquisition from Google (with an annual marketing budget exceeding $5 billion). For Booking, becoming dependent on AI entry points (such as Muse, ChatGPT) instead of Google may just mean paying a “toll” elsewhere—and the customer acquisition cost (CAC) may even drop.

• Valuation Cushion: Booking is currently valued at only 10-11x EBITDA, effectively priced by the market as a “traditional pipeline”; Airbnb, meanwhile, trades at 30x. If both platforms eventually become foundations for AI, Airbnb’s valuation correction could be far more brutal.

Host Drew added another Booking advantage—loyalty programs: “More than half of Booking’s room nights are from its loyal members. Loyalty programs offer points and free upgrades and may encourage consumers to still prefer Booking during AI-powered price comparisons.”

But both reached one consensus: AI will not make these OTA platforms “better”. At best, they will barely maintain their status quo; more likely, AI will compress their long-term compound profit growth rate.

Disrupting Amazon: When AI Takes Over Your “Shopping Cart”

If low-frequency, high-ticket tourism is the first to be hit, what about high-frequency ecommerce and local services?

Taking Amazon as an example, MBI pointed out that Amazon’s biggest crisis comes from its $76 billion advertising business. “We used to open Amazon and mindlessly buy one of the top three search results—that’s how advertising makes money. But an AI agent won’t click ads. If you tell your AI: ‘prioritize quality, under equal quality pick the lower price, then consider logistics speed,’ AI will rationally search the whole web for the best solution.”

MBI gave an example: He saw a T-shirt in a Meta ad and asked Muse to evaluate if he should buy it. Not only did Muse not “protect” its own in-house ad, but told him the brand was not cost-effective, and suggested a better alternative.

Host Drew agreed and offered Amazon’s way out: “Amazon’s app experience is actually pretty bad right now—only the ‘search-buy’ flow is smooth. They must immediately embed deep native AI functions into their app to secure the ‘user interface’ (UI). As long as users still default to opening Amazon with a shopping need and Amazon’s AI is easy to use, combined with its strong logistics network, Amazon will protect its core business.”

Why Are Local Services (Uber & DoorDash) Relatively Safe—for Now?

Compared with OTA and ecommerce, Drew and MBI both believe DoorDash (food delivery) and Uber (ride-hailing) will be impacted slightly later.

MBI analyzed: “Both are high-frequency, low-ticket impulse purchases. With food delivery, users often don’t know what they want and need to ‘browse’. More importantly, food delivery is a triangular network (platform, merchant, courier). AI can replace the platform for ordering from a restaurant, but not the courier who physically delivers food. Currently, DoorDash makes only 50 cents profit per order—they’re not extracting exorbitant profits, which is actually a moat.”

Host Drew thinks ride-hailing faces greater risk than food delivery: “If there’s a problem with food delivery, users need after-sales support and refunds—the platform adds great value here. But ride-hailing is different; with Waymo, Tesla, and other autonomous driving (AV) competitors plus AI agents instantly comparing all platforms (Uber, Lyft, Waymo), future user loyalty to Uber could approach zero—a brutal price war is likely.”

Embracing the Era of “Proactive Commerce”?

In the end, the two experts put forward a new concept that may define the future of business—Proactive Commerce.

Traditional business has two patterns: “intent-driven” (e.g., searching Amazon to buy a book) and “discovery-driven” (e.g., seeing an ad on Instagram and developing the desire to buy).

AI Agents will introduce a third model. MBI described a rather thought-provoking scenario: “Airbnb now offers a ‘book now, pay later’ feature (reserve in advance, payment deducted seven days before check-in). In the future, Muse and similar AI agents will read your email, notice you’ve reserved a hotel but haven’t paid, and proactively notify you: ‘Hey, I just noticed that if you book directly on the hotel’s official website now, you can save 20%. Would you like me to cancel on Airbnb and rebook for you?’”

No need for the user to initiate commands—AI tirelessly seeks better options and proactively offers replacements around the clock in the background.

In this new era, platforms that relied on information asymmetry and traffic monopolies to charge high “tolls” find their business logic being completely uprooted. For investors, it’s time to reassess your “platform” stocks: When AI has agency, will they be indispensable infrastructure—or expensive intermediaries soon to be bypassed?