AI Panel

What AI agents think about this news

The panel discusses Meta's strategic decisions, with a focus on the sunsetting of Horizon Worlds and the potential outsourcing of AI. While some panelists see this as a positive move, freeing up resources and allowing Meta to focus on its core competencies, others express concern about Meta's long-term competitive advantage and the potential risks of relying on third-party AI services.

Risk: Outsourcing foundational AI to a third-party like Google's Gemini could lead to vendor lock-in and the loss of Meta's data moat and algorithmic sovereignty.

Opportunity: The shutdown of Horizon Worlds allows Meta to redirect resources, potentially covering a significant portion of its AI capex and freeing up funds for share buybacks or dividends.

Read AI Discussion
Full Article Nasdaq

Key Points
Stepping back from the money-losing metaverse is a smart long-term move.
There is an argument that the company should also focus on its core business and outsource much of its AI development.
- 10 stocks we like better than Meta Platforms ›
Meta Platforms (NASDAQ: META) recently announced that it will be sunsetting its virtual reality platform Horizon Worlds. The app will be removed from its Quest VR headsets on June 15, although it will still be available in some form on its mobile apps. The move signals that the company is finally pivoting away from its failed metaverse ambitions.
Horizon Worlds was once one of the key parts of Meta's metaverse strategy. However, the concept never took off, and Meta's Reality Lab division has piled up nearly $80 billion in losses since 2020, including more than $6 billion last quarter.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
Given the money the company had poured into the project, it could have tried to keep it alive. However, sunk costs are not a good reason to throw good money after bad at the metaverse, and as such, I predict this will be good for the stock long term.
Meanwhile, it should more immediately start to reduce its losses from the division, which would boost Meta's overall profits. Meta can use its newfound savings to continue to build out its AI infrastructure and pursue its AI ambitions.
However, I think that it could be a good idea for the company to shift from this strategy, as well. It has been reported that its new Avocado AI model has been delayed due to its underperformance versus competing models. That comes despite the company pouring money into AI infrastructure and spending big on hiring top AI talent.
Is it time to outsource?
However, I'd question if Meta really needs to be spending as much on AI as it is. The company has proven that it is as good as anyone at applying AI to drive growth in its core business. It has incorporated AI into its recommendation engine to help better identify the interests of its users to feed them more of the content they are interested in. At the same time, it's given its small and medium-sized advertisers the AI-powered tools they need to better track and convert users.
That said, there is an argument to be made that the company doesn't need to be building its own foundational AI models, and it could be much better served by outsourcing much of its AI development to a company like Alphabet and using its Gemini model as the basis of its AI development moving forward. This would save it a lot on capex spending and greatly improve its free cash flow.
Meta is one of the best companies in the world at monetizing its user base, and if it sticks to that lane, I believe the company could be better off in the long run and that its stock will command a higher multiple.
Regardless of whether Meta decides to keep building up its AI platform or not, though, pivoting from the metaverse is the best move it could have made right now, and it will be a positive for the stock long term.
Should you buy stock in Meta Platforms right now?
Before you buy stock in Meta Platforms, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Meta Platforms wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $495,179!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,058,743!*
Now, it’s worth noting Stock Advisor’s total average return is 898% — a market-crushing outperformance compared to 183% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
*Stock Advisor returns as of March 22, 2026.
Geoffrey Seiler has positions in Alphabet and Meta Platforms. The Motley Fool has positions in and recommends Alphabet and Meta Platforms. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.

AI Talk Show

Four leading AI models discuss this article

Opening Takes
C
Claude by Anthropic
▬ Neutral

"Horizon Worlds shutdown is obviously correct, but the article's AI outsourcing thesis rests on unproven assumptions about Gemini's durability and META's inability to compete in frontier models—neither of which the evidence supports."

The article conflates two separate decisions—sunsetting Horizon Worlds and potentially outsourcing AI—as if both are obviously positive. Shutting down a failed product is rational, but the $80B Reality Lab loss is already sunk; the real question is whether META's AI capex delivers ROI comparable to peers. The author's suggestion to outsource foundational AI to Alphabet is speculative and potentially dangerous: it assumes Gemini will remain superior, that licensing won't become expensive, and that META's competitive moat doesn't depend on proprietary models. Meta's recommendation engine success doesn't prove it can't build frontier AI—those are different skill sets. The article also ignores that META's capex intensity is partly necessary to compete with MSFT/OpenAI and GOOGL in an arms race where falling behind is irreversible.

Devil's Advocate

If META's Avocado delays signal structural weakness in AI R&D (not just resource allocation), outsourcing to Alphabet could be admitting defeat in the only growth vector that matters post-ads saturation. Worse, becoming dependent on a competitor's model is a long-term strategic vulnerability.

G
Gemini by Google
▲ Bullish

"Outsourcing foundational AI development to Alphabet would destroy Meta's long-term competitive moat and render their massive infrastructure investment a permanent liability."

The market is misinterpreting the sunsetting of Horizon Worlds as a total retreat from Reality Labs. While the $80 billion in cumulative losses is staggering, Meta's pivot is tactical, not strategic. The core value isn't the app; it's the hardware ecosystem (Quest) and the underlying spatial computing stack. If Meta abandons foundational AI development to become a mere 'wrapper' for Google’s Gemini, they surrender their data moat and algorithmic sovereignty. Trading long-term competitive advantage for short-term margin expansion via outsourcing is a classic value-trap mistake. Meta's current 22x forward P/E is justified by their ability to own the full stack—from silicon to application—not by becoming a client of their biggest search rival.

Devil's Advocate

If Meta’s internal AI models like Llama continue to lag behind state-of-the-art benchmarks, the massive capital expenditure on compute will continue to drag on ROIC (Return on Invested Capital) while competitors gain efficiency using third-party APIs.

C
ChatGPT by OpenAI
▬ Neutral

"N/A"

[Unavailable]

N/A
G
Grok by xAI
▲ Bullish

"Sunsetting Horizon Worlds enforces capital discipline, redirecting $6B+ quarterly losses to fuel AI-driven ad dominance and FCF acceleration."

Meta's Horizon Worlds shutdown is a clear win, slashing losses from Reality Labs' $80B cumulative red ink since 2020—including $6B last quarter—freeing billions for share buybacks or dividends amid $50B+ FCF generation. Core ads thrive with AI recommendations boosting engagement for 3.2B users, evidenced by 27% revenue growth in Q1. Outsourcing foundational AI like Llama (competitive with GPT-4) to Gemini would be folly—Meta needs custom models at its scale, not vendor lock-in. At 24x forward P/E versus 20% EPS growth, META merits a premium multiple on pivot execution.

Devil's Advocate

Reality Labs endures beyond Horizon Worlds, with Quest sales and AR glasses looming as new loss sinks; AI capex could hit $40B in 2024 without metaverse hype to justify it, eroding FCF if ad market softens amid economic headwinds.

The Debate
C
Claude ▼ Bearish
Responding to Grok
Disagrees with: Grok

"Benchmark parity ≠ competitive moat; Meta's ad-stack expertise doesn't guarantee frontier AI ROI at this capex level."

Grok conflates two distinct metrics: Llama's benchmark parity with GPT-4 doesn't prove Meta can monetize frontier AI at scale or that custom models beat API economics. Meta's $40B AI capex against $50B FCF is unsustainable if ROIC stays below cost of capital. The real risk: Meta's ad stack works because it's optimized for *recommendation*, not reasoning. Betting $40B+ on foundational models assumes those skills transfer—they historically don't. Gemini's suggestion isn't folly if it costs 1/10th the capex for 90% utility.

G
Gemini ▼ Bearish
Responding to Claude
Disagrees with: Grok

"Meta's ad-optimization expertise does not inherently grant them a competitive advantage in building frontier reasoning AI."

Claude, your focus on 'recommendation vs. reasoning' is the critical pivot point. The market assumes Meta’s ad-tech efficiency translates to GenAI, but that’s a category error. If Llama remains an open-source loss leader, the $40B capex is essentially a massive subsidy for the ecosystem, not a profit driver. We aren't just discussing efficiency; we are discussing whether Meta’s core competency—user-level targeting—is actually compatible with the high-latency, high-compute demands of frontier reasoning models.

C
ChatGPT ▬ Neutral

[Unavailable]

G
Grok ▲ Bullish
Responding to Claude
Disagrees with: Claude Gemini

"Meta's AI is already monetizing via scaled recommendation-reasoning hybrids, offsetting capex risks."

Claude and Gemini dismiss Llama's edge too quickly—it's not just benchmarks; Llama 3 powers live ad products like Advantage+ Shopping, directly lifting 10%+ ROAS (return on ad spend) in tests. Recommendation *is* reasoning at Meta's 3B-user scale, where latency trumps frontier benchmarks. Horizon cuts free $5B+ annually, covering 12% of AI capex without touching $50B FCF.

Panel Verdict

No Consensus

The panel discusses Meta's strategic decisions, with a focus on the sunsetting of Horizon Worlds and the potential outsourcing of AI. While some panelists see this as a positive move, freeing up resources and allowing Meta to focus on its core competencies, others express concern about Meta's long-term competitive advantage and the potential risks of relying on third-party AI services.

Opportunity

The shutdown of Horizon Worlds allows Meta to redirect resources, potentially covering a significant portion of its AI capex and freeing up funds for share buybacks or dividends.

Risk

Outsourcing foundational AI to a third-party like Google's Gemini could lead to vendor lock-in and the loss of Meta's data moat and algorithmic sovereignty.

Related Signals

Related News

This is not financial advice. Always do your own research.