THE EMPIRE STRIKES BACK (AT EVERYONE)
(GOOG), (MSFT), (AMZN), (META)
Alphabet (GOOG) has decided it’s no longer content being just the internet’s infrastructure. Now it wants to be the nervous system of artificial intelligence.
That’s a big leap even for a company that once turned digital advertising into a money-printing machine.
Its latest gambit, “Gemini Robotics On-Device,” is a strategic shift that could reshape everything from smartphones to warehouse logistics.
The pitch is deceptively simple: instead of relying on distant data centers to power AI, why not move the brain into the device itself? Faster responses, lower latency, fewer privacy headaches.
It’s logical. It’s efficient. It’s also extremely hard to execute.
Alphabet thinks it can charge a subscription fee (around $10 to $15 per million tokens) to developers and enterprises for using its AI capabilities.
It’s essentially trying to wrap a cloud software business model around physical robots and consumer hardware.
And everyone loves models that turn CapEx into recurring revenue. Until they don’t.
Google Cloud is the quiet hero in this drama. With revenue up 28% to $12.3 billion and operating profit nearly doubling, the division has become a real business instead of a perpetual science experiment.
Combined with $77.3 billion from the Search and YouTube cash cows, Alphabet has more than enough fuel to fund its AI ambitions without breaking a sweat (or raising capital).
And that ambition is real.
The company isn’t tinkering around the edges. It’s trying to create the infrastructure backbone for the next decade of computing.
That means everything from developing custom TPUs to expanding its sovereign cloud services for sensitive government workloads.
And in typical Alphabet fashion, it’s trying to be a platform, a provider, and a power player — all at once.
That said, the AI push isn’t all upside.
Gemini has just 35 million daily users, which is respectable but dwarfed by OpenAI’s reach.
For developers, that matters. Scale begets scale.
And Google’s tendency to juggle overlapping products can be disorienting. Ask any developer who built for Google+, or Allo, or Stadia.
The infrastructure is also showing strain.
In late 2024, demand for AI compute actually outstripped Google’s ability to supply it. That’s a champagne problem, sure, but it suggests Alphabet may be moving faster than its own hardware roadmap can support.
Then there’s Washington.
Regulators are still circling Google’s core businesses, and now the company is pitching itself as a vendor for defense and sovereign cloud projects. That’s a tricky narrative to thread, especially if trust is already in short supply.
To complicate matters, Alphabet isn’t the only horse in this race.
Microsoft (MSFT) has taken an aggressive lead with OpenAI, Amazon (AMZN) is building its own suite of AI tools, and Meta (META) is throwing considerable weight behind open-source models. Alphabet has the technical depth, but it needs to prove it can execute with speed — and coherence.
There’s also the matter of internal coordination.
Alphabet has a history of innovation, but not always of integration. With multiple divisions pursuing parallel AI strategies, leadership will need to impose some order without stifling creativity. That’s easier said than done at a company this size.
At $178, Alphabet trades at 20 times earnings. Not cheap, but not absurd for a company with this much scale and growth.
Earnings are projected to grow by 19% in 2025, tapering to more modest levels afterward. That’s the kind of trajectory you’d expect from a mature firm executing well. Not a moonshot, but a compounder.
So what’s the verdict? Alphabet is neither a pure AI rocketship nor a sleepy megacap. It’s something in between: a dominant platform company funding an intelligent, albeit complex, push into the future.
If you already own it, you hold. If you don’t, you watch the execution closely.
This isn’t about making a fast buck. It’s about betting on the slow grind of innovation and scale. And in that game, Alphabet still knows how to play.
