To solve long-term constraints like land and power, Google CEO Sundar Pichai revealed the company is exploring a new moonshot project: data centers in space. While in the very early stages, it represents the kind of thinking required to sustain AI's growth over a multi-decade horizon.

An innovation arm's performance isn't its "batting average." If a team pursues truly ambitious, "exotic" opportunities, a high failure rate is an expected and even positive signal. An overly high success rate suggests the team is only taking safe, incremental bets, defeating its purpose.

Google's "Project Suncatcher" aims to place AI data centers in orbit for efficient solar power. However, the project's viability isn't just a technical challenge; it fundamentally requires space transport costs to decrease tenfold. This massive economic hurdle, more than technical feasibility, defines it as a long-term "moonshot" initiative.

The default assumption for any 'moonshot' idea is that it is likely wrong. The team's immediate goal is to find the fatal flaw as fast as possible. This counterintuitive approach avoids emotional attachment and speeds up the overall innovation cycle by prioritizing learning over being right.

While experts dismiss Elon Musk's idea of space-based AI data centers as unviable, this overlooks his history with SpaceX, which consistently achieves what was deemed impossible, like reusable rockets. His analysis of the physics and economics may be more advanced than public criticism allows.

Skepticism around orbital data centers mirrors early doubts about Starlink, which was initially deemed economically unfeasible. However, SpaceX drastically reduced satellite launch costs by 20x, turning a "pipe dream" into a valuable business. This precedent suggests a similar path to viability exists for space-based AI compute.

SpaceX manages its aggressive "fail fast" culture by creating distinct risk profiles. Development projects like Starship are intentionally pushed to failure for learning. In contrast, operational, human-rated systems like Dragon are built with massive safety margins and exhaustive, conservative testing.

For ambitious 'moonshot' projects, the vast majority of time and effort (90%) is spent on learning, exploration, and discovering the right thing to build. The actual construction is a small fraction (10%) of the total work. This reframes failure as a critical and expected part of the learning process.

In a new technological wave like AI, a high project failure rate is desirable. It indicates that a company is aggressively experimenting and pushing boundaries to discover what provides real value, rather than being too conservative.