Astro Teller, head of Alphabet’s Moonshot Factory X, during an interview with DLD, sitting in a video studio, speaking in front of a blue screen showing the DLD Munich 2025 design: icons showing yellow hearts and red globes.

DLD backstage interview

Radical Innovation: “I’m Not Giving You Money So You Can Succeed”

Astro Teller, head of Alphabet’s Moonshot Factory, tells DLD why world-changing ideas world must beat the odds of failure – again and again, in the most efficient way.

Karsten Lemm

February 28, 2025

DLD25

Astro Teller is beaming with pride as he tells the story of a belly flop. One of his teams recently came to him with an hours-long presentation, explaining why their project should be shut down. Years of work: gone. High-flying hopes for an idea that could change the world and become a successful business: shattered.

And yet, that’s exactly the way things are supposed to work at X – The Moonshot Factory, Alphabet’s in-house innovation lab, which Teller has been leading as “Captain of Moonshots” since its inception in 2010, a time when the notoriously secretive lab was still known as Google X.

“Our focus is on particularly audacious things that would be really good for the world and could turn into enduring businesses if we succeed”, Teller says.

Try. Fail, Fail, Fail. Try Again!

Google Brain, which helped lay the foundation for the current AI boom; Waymo, a pioneer of self-driving cars; and Alphabet’s digital health unit Verily all came out of X.

But for every high-profile success, there are dozens and dozens of failures – such as the recent moonshot that fell to earth. “It was the right thing to do”, Teller says. “The team suggested it. We’re super proud of them for the work they did. Super proud of them for ending their own project.”

Captain of Moonshots

Eric “Astro” Teller studied computer science at Stanford and Carnegie Mellon universities. Before joining Google X in 2010, he founded several companies, including BodyMedia, a wearable body monitoring company, and Cerebellum Capital, a hedge fund management firm utilizing statistical machine learning. He is the grandson of the famous physicist Edward Teller.

Speaker Profile

By Teller’s count, X has evaluated some 2,000 potential moonshots over the past 15 years. Always with a clear focus on three basic elements that need to come together for any idea to get a chance.

First, “there has to be a huge problem with the world that you want to solve”, Teller explains. Otherwise “there’s a real danger that the whole effort might be academic and not actually meaningful.”

The second ingredient is a “science-fiction sounding product or service” that would solve the problem – but it can’t just be a fanciful notion, Teller emphasizes. “There has to be some kind of breakthrough technology that you can name” which might turn the idea into a reality.

Audacity Meets Efficiency

Once a proposed moonshot passes muster, Alphabet is willing to give it a try. “Of course, we couldn’t afford to try all of them at scale”, Teller says. That’s why budgets are low, and teams are expected to kill their ideas as fast as possible.

Astro Teller on radical innovation: Watch our DLD25 backstage interview.

“We’re constantly saying, ‘This looks exciting – and it’s probably wrong’”, Teller explains. “What is the cheapest way, what is the fastest way we can verify that this is not one of those once-in-a-generation opportunities. So we can, with love, put it away and move on to the next one.”

The idea behind the process is to marry audacity with efficiency, Teller told WIRED editor Greg Williams at DLD Munich 2025. “It’s really easy to take moonshots if you don’t care about efficiency”, Teller said. “You just take a huge amount of money, you give it to some really smart, high-energy people, and you will get a few cool things. But you will be sorry about how you spent your money. It won’t be efficient.” (Video below.)

Alphabet, by contrast, aims to be as frugal as possible while trying to figure out whether an idea is worth pursuing further. “We are obsessed with the reward-risk ratio“, Teller explained at DLD25. “It mostly is about, ‘I’m not giving you this money so you can succeed.’ Because you’re trying something which has a 1% chance of succeeding. And so I want to stop the 99% that are wrong as fast as possible.”

Limitation Breeds Creativity

All the while, innovators at X are pursuing their ideas on a minimal budget. They may get $10,000 or $30,000 to take their project to the next level – assuming it survives – but they’re not swimming in money, even though Alphabet could easily afford it. In 2024, Google’s mother company posted a net profit of $100 billion on revenues of $350 billion.

“We are purposefully trying to keep the teams at X very tight”, Teller says, because limitation, Alphabet hopes, will breed creativity. “Under-resourcing them is a strategy at X, so that the only way they could succeed would be to find some breakthrough that’s so much better than the alternatives, that even being under-resourced they can still succeed.”

Patience, Please!

Is the concept working? Yes, says Teller, “or I wouldn’t keep doing it.” Similarly, Alphabet “wouldn’t keep paying for it”, if the company didn’t see real value in each active project on the horizon.

But radical innovation is not for the impatient, because there aren’t any direct benefits for next quarter’s bottom line. “The time from having a crazy idea until it is so obviously a good idea, and a profitable thing, that no one is discussing whether it’s a good idea anymore – that’s often 10 to as much as 20 years”, Teller notes.

“You just have to sign up for that”, he adds, “or you shouldn’t be in the innovation business.”

WIRED editor Greg Williams (left) and Astro Teller, head of Alphabet’s Moonshot Factory, speak at the DLD Munich 2025 conference. Video preview of their talk.

Challenge Impossible

Watch the DLD25 conversation between Astro Teller and WIRED’s Greg Williams about “Lessons Learned  from 15 Years of Moonshot Taking”.

Watch

By loading the video you agree to the privacy policy of Youtube.

Currently, a moonshot called Tapestry gives him hope that his factory has produced another hit. Tapestry uses AI and data analytics to help energy companies manage their electric grids more intelligently.

“It’s seven years old from our perspective, so we feel like it’s old news”, Teller says. “But I predict in another seven years it will be as well known as Waymo.”

Looking ahead, the Captain of Moonshots is confident that radical innovation can help solve some of the world’s biggest challenges – including better health care, education, and mitigating the effects of climate change – and that his teams will play a role in making it happen.

“There are lots of things that are still within X, or recently have left X, that are as exciting as the self-driving cars or Google Brain or Wing”, Teller says. “It’s just because they’ve had seven or eight years less runtime that they’re not as known by the world yet.”

And looking at the past 15 years, he adds, “I can say that the winnowing that we’re doing, the value of the things that we’re getting, the goodness that they’re doing for the world – yes, I think it’s working.”