Google just sent AI chips into orbit.
At roughly the same time, a federal judge handed the company a significant victory over publishers challenging the way Google uses AI-generated answers in search.
The stories appear unrelated. One involves satellites, solar power and the enormous computing requirements behind artificial intelligence. The other involves publishers, search traffic and the economics of being discovered online.
For revenue leaders, however, they point in the same direction.
The infrastructure surrounding AI is changing at both ends of the equation: where intelligence is computed and where customers encounter it.
And companies building their go-to-market strategies around yesterday’s infrastructure may have more to rethink than they realize.
Google’s Data Center Just Left the Planet
The long-term idea sounds almost like science fiction: put machine-learning infrastructure in space.
Project Suncatcher explores whether interconnected, solar-powered satellites equipped with Google Tensor Processing Units could eventually provide large-scale AI computing capacity beyond terrestrial data centers. Google says solar panels positioned in the right orbit can be up to eight times more productive than panels on Earth and generate power almost continuously.
The prototype is considerably more modest.
Google is testing how its TPUs perform through the physical stresses of launch and the radiation and thermal extremes of orbit. The experiment could help determine whether the economics and engineering of orbital computing are remotely practical at scale.
There are substantial hurdles.
Cooling high-powered chips in space is difficult. Launch costs remain high. Future computing clusters would require extremely fast communication between satellites. Google itself describes Suncatcher as a long-term research moonshot, not an imminent replacement for terrestrial data centers.
But the direction is noteworthy.
The AI infrastructure race is becoming an energy race, a compute race and increasingly an infrastructure-design race.
For CFOs and technology leaders trying to forecast the economics of AI, that matters.
The question is no longer simply how much cloud capacity an organization needs. Executives increasingly need to understand how changes in chips, energy generation, cooling, networking and alternative computing infrastructure could eventually affect the unit economics of AI.
Google is literally looking beyond Earth for additional possibilities.
Meanwhile, Google Is Changing What “Search” Means
While Google experiments with where AI computing happens, another battle is unfolding over where information gets consumed.
Chegg and Penske Media had challenged Google’s AI Overviews under antitrust law, arguing, among other things, that Google’s AI-generated summaries use publisher content while reducing the need for users to visit the original websites.
On September 30, U.S. District Judge Amit Mehta dismissed the companies’ antitrust claims.
One part of the ruling gets directly to the changing economics of search: publishers may have historically expected Google to send them traffic in return for making their content available for indexing, but the court concluded that an expectation of traffic did not amount to the kind of agreement required for the plaintiffs’ theory.
For more than two decades, digital marketing operated around a relatively straightforward transaction: Get discovered. Get the click. Get the visitor. Convert the visitor. AI search is breaking that sequence apart.
A buyer can increasingly ask a question, receive a synthesized response and continue researching without ever visiting the websites whose information helped produce that answer. The click is no longer guaranteed to be the moment when your company enters the buying journey.
The New Marketing Question: Are You in the Answer?
This creates a different challenge for CMOs.
Traditional SEO asks if buyers can find us? AI-driven discovery introduces another question: Does the answer engine know enough about us to include us?
Those aren’t identical problems.
Companies still need technically sound websites, authoritative content and strong organic visibility. But increasingly, they also need information that answer engines can understand, retrieve, corroborate and cite. That puts greater value on clear expertise, original research, structured information, third-party validation and consistent descriptions of what a company actually does.
In other words, traffic may no longer be the only meaningful measure of visibility. Citation share, answer inclusion and brand presence inside AI-generated research experiences are becoming part of the equation.
And That Changes the Revenue Conversation
This isn’t only a marketing issue. When buyers arrive at a sales conversation after conducting substantial research through AI assistants, the top of the funnel begins happening somewhere the seller can’t see.
The buyer may already have asked:
- Which platforms solve this problem?
- How do these vendors compare?
- What does implementation typically cost?
- What problems should I expect?
- Which capabilities actually matter?
That research may happen before a form fill, demo request or identifiable website visit. For CROs and RevOps leaders, that creates a measurement problem. Your CRM may show when a prospect entered your pipeline. It may not show when the prospect first encountered your company. And that distinction becomes increasingly important as AI intermediates more of the buying journey.
Your GTM Infrastructure Has to Catch Up
Google is reconsidering the physical infrastructure required to run AI. Meanwhile, marketers are reconsidering the information infrastructure required to be discovered by AI.
Revenue organizations should be asking, “Is our go-to-market infrastructure designed for the way buyers actually make decisions now?
That includes far more than generating additional content or adding another AI tool. Revenue teams still have to answer fundamental operational questions:
- Who owns this account?
- Which prospects match our ICP?
- How should leads be routed?
- Where is pipeline deteriorating?
- Are territories aligned with market opportunity?
- Are quotas achievable?
- Which deals require intervention?
- Are compensation plans reinforcing the behavior the business actually wants?
Those decisions require something AI alone cannot manufacture: a coherent revenue plan.
AI Can Change the Interface. It Can’t Fix a Broken Revenue Model.
The temptation with every major AI development is to focus on the technology itself.
Project Suncatcher certainly deserves attention. Putting TPUs into orbit makes for a spectacular headline. But the more consequential lesson for executives may be much less exotic. AI is changing the infrastructure businesses rely on.
Compute infrastructure is changing. Search infrastructure is changing. Buyer research is changing. And the systems revenue organizations use to plan, execute and measure their GTM motions will have to change with them. That’s why the next competitive advantage may not come from adopting one more AI application.
It may come from making sure AI has something reliable to work with in the first place: clearly defined territories, accurate capacity plans, realistic quotas, intelligent routing, measurable performance signals, reliable forecasting and compensation aligned with the revenue strategy.
The companies that solve that operational layer will be better positioned no matter where the computing happens. Even if the data center happens to be orbiting 400 miles above their heads.
________________________________
4 KEY TAKEAWAYS
4 Key Takeaways
1. Why is Google putting AI infrastructure in space?
Project Suncatcher is testing whether solar-powered satellites equipped with TPUs could eventually provide another source of large-scale AI compute, potentially reducing some terrestrial energy and infrastructure constraints.
2. How is AI changing search and brand discovery?
AI-generated answers increasingly give buyers information without requiring a website visit, making brand inclusion and citation visibility more important alongside traditional search rankings.
3. What does zero-click discovery mean for revenue teams?
Buyers can research vendors, compare solutions and form opinions before appearing in a CRM or submitting a form. Revenue teams need to account for an increasingly invisible early buying journey.
4. What should RevOps leaders take from Google’s AI moves?
As AI reshapes compute, search and buyer behavior, GTM infrastructure needs to keep pace. Clean territories, accurate quotas, intelligent routing, reliable forecasting and aligned compensation give AI—and revenue teams—a stronger operational foundation.