For most of this year, the market treated Meta like it had finally cracked the AI code.

Revenue up 33%. Ad prices rising. The Llama models actually working inside the ad stack. Every quarter seemed to prove the skeptics wrong.

Then Zuckerberg raised capex guidance to $145 billion for 2026, nearly double what Meta spent in all of 2025, and admitted on the earnings call that he did not have a "very precise plan" for how each AI product would scale.

The stock fell 6% in after-hours. It is now sitting 27% below its August 2025 all-time high.

And honestly, I think that admission matters far more than most investors want to sit with.

Because the question is no longer whether AI is working at Meta. It clearly is, inside the ad business. The question is what justifies the other $100 billion sitting on a plan that has not been written yet.

Today in One Big Idea, I am pulling that question apart.

The real point here is that SpaceX is no longer just making Elon Musk richer. It is creating a wider circle of ultra-wealthy winners as the post-IPO valuation surge lifts major early backers and insiders along with him. Reuters reported last week that Musk’s net worth had already crossed about $1.1 trillion after the IPO, with most of that wealth tied to SpaceX.

The Fool’s argument is that the EV trade has simply gone cold, not dead. It points to a still-fast-growing global EV market and highlights Rivian and Nio as beaten-down names that could benefit if sentiment recovers, with Rivian pitching the R2 as its turnaround catalyst and Nio leaning on delivery growth, battery swapping, and improving margins.

This MarketBeat piece makes a solid case that the Google Cloud partnership is really about distribution. Palantir just posted 85% year-over-year revenue growth to $1.63 billion, then added Google Cloud to a cloud roster that already includes AWS, Azure, and Oracle, which reduces adoption friction by letting customers buy and deploy Palantir inside the infrastructure they already use. In plain English, the deal does not magically solve Palantir’s valuation debate, but it does make the company look more embedded in the enterprise AI stack.

58% of buyers now start their research in ChatGPT or Gemini, not Google. Most startups aren't showing up there yet.

The ones that are get cited by the AI tools their buyers, investors, and future hires already use. And they convert at 3×.

Download the free AEO Playbook for Startups from HubSpot and get the exact steps to start showing up. Five minutes to read.

Get the free playbook

The obvious read on Meta right now is strong. Revenue up 33%, the fastest quarterly growth since 2021. Ad revenue at $55 billion. Price per ad up 12% year over year. The AI-powered ad machine is running at full throttle.

That read is not wrong. It is just incomplete.

Here is the part that matters. Zuckerberg raised 2026 capex guidance to $125 billion to $145 billion at the same time those numbers landed. In 2025, Meta spent $72.2 billion on capital expenditure. They are now guiding to nearly double that figure.

And when asked how each AI product would generate returns, the CEO said the company does not yet have a "very precise plan." That is the question One Big Idea is here to examine this week.

Let me be direct about something before I get to the risks. Meta's advertising business is extraordinary, and the numbers make that difficult to dispute.

Q1 2026 revenue hit $56.31 billion, up 33% year over year, ahead of the $55.45 billion estimate. Ad revenue alone came in at $55.02 billion.

Price per ad rose 12% year on year. For the full year 2026, Meta's ad revenue is projected to reach $240 billion, 22.3% growth on an already enormous base.

That growth is not happening despite AI. It is happening because of it.

Meta's Andromeda ad delivery system and deep Llama 4 model integration have made the ad auction significantly smarter. Better targeting means higher relevance, better relevance means better advertiser outcomes, and better outcomes mean advertisers pay more per impression.

The 12% price per ad increase in Q1 is the clearest evidence this cycle is working.

Because when AI makes your core product measurably better and advertisers respond by paying 12% more for it, that is not a story about the future. That is revenue happening right now.

That is what the bull case is built on. And it is a genuinely strong foundation.

Here is the part most Meta bulls are not sitting with long enough.

Meta is spending $145 billion in a single year on capital expenditure. That is more than what Meta spent in 2024 and 2025 combined. Unlike Google, Amazon, and Microsoft, Meta has no cloud business to directly monetise AI infrastructure by renting compute to enterprise customers.

Nearly 98% of Meta's revenue comes from advertising.

That means every dollar of that $145 billion must be justified by either growing the ad business further or creating a new revenue stream that does not currently exist at meaningful scale. Meta AI has no announced subscription model. Llama builds ecosystem positioning but no direct revenue.

Zuckerberg confirmed on the earnings call there is no "very precise plan" for how individual AI products will monetise.

Not a reason to sell tomorrow. A reason to ask the question seriously and size your position accordingly.

When a company doubles capex with 98% revenue concentration in one category and no clear second act, the market is right to price in real uncertainty.

That is what a move from $787 to $577 looks like.

META has spent ten months telling a story the headline revenue numbers do not fully capture.

The stock hit an all-time high of $787.42 in August 2025 and has been drifting lower since, sitting around $577 heading into this week. That is a 27% decline from peak in a period when AI-adjacent names have broadly held up. The selloff tracks almost exactly to when capex escalation became the dominant narrative and the ROI story started getting murky.

The pattern heading into this week is specific. Every rally toward $600 to $620 brings sellers. The Q1 earnings beat briefly sent the stock higher, but the capex guidance increase in the same report pulled it back. Buyers cannot hold conviction above $600 when the spend keeps rising and the payoff timeline keeps extending.

The level to watch is $540 to $550. That is where META found support during the last meaningful correction. A break below there signals institutional investors are genuinely reassessing the thesis, not waiting for a better entry.

Q1 was the opening data point. Q2 will either confirm the bull case or expose the first crack.

If ad revenue continues growing at 25% or more while capex guidance holds at the lower end of the range, the ROI narrative begins to repair. If revenue growth moderates while capex keeps climbing, the market's concern becomes structurally harder to dismiss with a single strong quarter.

Because the ratio of growth to spending matters more right now than the absolute revenue figure.

That is the metric I am watching before anything else in the next earnings release.

Meta AI claims over 1 billion monthly active users. It is currently free.

At some point, one billion users must translate into revenue beyond incremental advertising engagement. When Zuckerberg announces a direct monetisation pathway, whether subscription, premium tier, or enterprise licensing, that single announcement changes the thesis more than any quarterly earnings beat.

Not an assistant. A revenue line. Those are very different things.

That announcement would re-rate this stock faster and more durably than any capex reduction.

The 12% price per ad increase in Q1 2026 is the clearest evidence that AI targeting improvements are delivering real, measurable value at the auction level.

If that accelerates to 15% or higher in Q2, the AI gains inside the ad auction are compounding. If it decelerates toward 5% to 7%, it raises the serious question of whether the easy improvements have already been captured and further spending delivers diminishing returns.

Because price per ad is the purest measure of how much Meta's AI is actually converting into advertiser value, not just engagement metrics.

That is what I read before the revenue headline in every Meta report.

Zuckerberg has revised capex guidance upward more than once in 2026 already.

The market has tolerated each increase because ad revenue kept outpacing it. But there is a ceiling to what investors absorb without a clearer non-advertising payoff attached to the spending. If Q2 guidance pushes above $145 billion again without a corresponding product revenue announcement, META's ability to hold current levels gets seriously tested.

Because every upward revision without a new revenue model widens the gap between what is being spent and what is being earned. And that gap has a direct cost, measured in multiple compression.

That is the trust equation Zuckerberg is spending down, quarter by quarter.

I want to be direct about where I have landed on Meta.

The ad business is so strong it can absorb a significant amount of imprecise capital allocation without the core story breaking. A company generating $56 billion in quarterly revenue, growing at 33%, with price per ad up 12%, does not fall apart because of a capex number. That is not the risk.

The risk is that $145 billion of AI infrastructure gets built for a product ecosystem that never develops a revenue model beyond advertising. That risk is real precisely because Meta has no cloud business to fall back on, no enterprise compute revenue to offset the build costs, and a CEO who admitted on the record he cannot yet describe the precise plan.

Zuckerberg has been right before when the plan looked incomplete. The Year of Efficiency in 2023 was dismissed until the results came in. That history matters and I do not ignore it.

But the size of the bet this time is different. This is not a cost discipline story. This is a $145 billion infrastructure wager on a product ecosystem that has not yet shown how it generates the next leg of revenue.

The ad business is not the bet. The $145 billion sitting outside the ad business is the bet.

Know exactly which one you are owning before this week moves.

The next agricultural transition may not be bigger tractors. It may be autonomous robots replacing herbicides entirely. Greenfield Robotics is building commercial systems designed for that future.

Explore Early Access

_{Greenfield Robotics is Testing The Waters under tier 2 of Regulation A. No money or other consideration is being solicited, and if sent in response will not be accepted. No offer to buy the securities can be accepted and no part of the purchase price can be received until the offering statement filed by the company with the SEC has been qualified by the SEC. Any such offer may be withdrawn or revoked, without obligation or commitment of any kind, at any time before notice of acceptance given after the date of qualification. An indication of interest involves no obligation or commitment of any kind. “Reserving” shares is simply an indication of interest. There is no binding commitment for investors that reserve shares in this manner to ultimately invest and purchase the shares reserved of the company, or to purchase any shares of the company whatsoever.}

The most dangerous moment for any great business is when its own success starts funding bets that investors cannot evaluate.

Meta's ad machine generates so much cash it can afford to be imprecise. That is not the same as being right.

Zuckerberg admitted he does not have a precise plan for $145 billion in spending. That kind of honesty is rare.

It is also expensive.

The investors staying comfortable holding META here have decided to trust the operator, not the plan.

That is a different kind of conviction than the one that got the stock to $787.

Know which one you have before this week starts.