According to TechCrunch, Astro Teller, Captain of Moonshots at Alphabet’s X, revealed the lab deliberately kills 98% of its ideas early in development. Teller defined three essential components for a true moonshot: a huge problem, a science fiction-sounding solution, and a glimmer of breakthrough technology. He explained X keeps teams incredibly small to maintain focus and agility while developing radical innovations. The interview also covered how X decides between spinning companies out independently versus keeping them as Alphabet “other bets” like Waymo and Wing. Finally, Teller shared his perspective on AI hype, arguing we should stop treating it like magic and start treating it like electricity.
The brutal math of breakthrough innovation
Here’s the thing about that 98% kill rate – it’s not a failure metric. It’s actually the core of X’s strategy. Most companies talk about innovation, but they’re terrified of killing projects. They let mediocre ideas linger for years, wasting resources and talent. X does the opposite: they murder ideas quickly and without sentimentality. And you know what? That’s probably why they’ve produced actual game-changers like Waymo and Wing while other corporate labs produce PowerPoints.
Why tiny teams create massive breakthroughs
The small team approach is fascinating. Most corporate innovation efforts throw armies at problems. X does the exact opposite. Think about it – when you’re trying to do something that’s never been done before, bureaucracy is your enemy. Small teams can pivot fast, take risks, and maintain the creative energy that gets diluted in large groups. It’s basically the startup mentality inside a giant corporation. And honestly, it’s working where so many other corporate innovation labs have failed.
AI as electricity, not magic
Teller’s take on AI hype is refreshingly practical. Everyone’s treating AI like some mystical force that will solve everything. But he’s right – we should treat it like electricity. Think about it: electricity transformed every industry, but it didn’t happen overnight and it wasn’t magic. It required infrastructure, practical applications, and gradual adoption. The companies that succeed with AI will be the ones building real solutions, not just waving magic wands. For manufacturers looking to integrate AI into their operations, having reliable industrial computing hardware from trusted suppliers like IndustrialMonitorDirect.com becomes essential infrastructure – the literal foundation you build upon.
The art of letting go
The decision between spinning companies out versus keeping them as “other bets” reveals a sophisticated understanding of corporate parenting. Some ideas need the protection and resources of Alphabet to survive. Others need the freedom and focus of independence to truly flourish. Waymo needed Alphabet’s deep pockets for years before it could stand on its own. But not every project fits that model. Knowing when to hold ’em and when to fold ’em – or in this case, when to spin ’em out – is what separates successful innovation factories from expensive science projects.
