Something strange happened with Meta.
Mark Zuckerberg admitted AI agents were not progressing as quickly as executives expected.
The stock went up anyway.
That is the kind of market reaction that makes people stop and ask what exactly investors are pricing in.
Meta had spent months under pressure because investors were worried about the massive AI spending plan.
The company was expected to spend around $125 billion to $145 billion on AI infrastructure in 2026.
The big question was simple.
Where is the payoff?
Then Zuckerberg basically said the AI agent timeline was slower than expected and admitted the restructuring around AI had not gone as cleanly as planned.
You would expect that to scare investors.
Instead, the market found a different angle.
Some Reddit investors argued the stock was not rallying because AI was perfect.
They argued the real story was that Meta may have improved the economics behind the AI buildout, including lowering infrastructure costs and finding ways to monetize the massive computing investment.
That is the interesting part.
A few years ago, AI was sold as a race where everyone needed to build faster.
Now the conversation is slowly changing into another question.
Who can spend the least money and still win?
Some commenters pointed out Meta is still growing revenue strongly for a mature company and already has billions of users waiting for AI products.
Others were much less convinced.
They questioned whether a few good announcements before earnings are enough to prove that billions in AI spending will actually create real returns.
The funniest part is watching investor emotions flip.
When Meta was falling, people acted like the company was doomed.
After a rally, suddenly everyone is explaining why it was misunderstood.
Same company.
Same AI spending.
Same uncertainty.
Different stock price.
The bigger question is whether Meta has actually solved the AI problem or simply convinced investors that the problem is cheaper than expected.
Because the market can forgive delays.
It can forgive huge spending.
But eventually it still needs results.
The next earnings report will show whether this AI comeback is backed by numbers or just another round of optimism.