"What’s Going On At Google[X]" in Humanizing Tech
An update on the Moonshot factory
I. Why Even Bet on X?
Alphabet reported its Q2 2016 earnings last week. It was almost entirely positive except for two things: Google’s advertising price per click is decreasing and Other Bets are losing nearly $1 billion. Alphabet doesn’t break out the Other Bets into operating units like the Google[X] Moonshot factory, but it matters not. Most new startups like these inside Fortune 500 corporations take a lot of R&D before they pay off. And even then, the probability of success is maybe 1 in 10.
Think about where Google makes all of its money. For the last few decades they invested in countless projects, products, apps, devices, and operating systems. But to this day, 90% of Alphabet’s revenue comes from companies paying to advertise their stuff right next to your Google search. Even YouTube, at their massive scale, is basically a break-even business.
It begs the question of why any company would continue to keep trying decade after decade, year after year, day after, engineer after engineer when nothing else has seemed to pay off? Would they have been better off hoarding their cash, repurchasing their own stock, and eventually taking themselves private? I’m not so sure.
I believe the reason they keep going is two fold:
- Keeps people engaged. Google still needs to be relevant and top-of-mind for when you search for something. It’s their cash cow. They need to protect it with a massive moat and big ole drawbridge. Google also still needs to attract investors looking for growth. And they still need to attract high-caliber talent in order to keep building all the things.
- It only takes one grand slam. Apple wouldn’t be the Apple we know today without the iPhone. It single handedly made the company and gave it the cash hoard it has today. Google is searching for, and will eventually, get their iPhone. It may just take another 10 years. It might be AI. It might be AR/VR. It might be a differentiated self-driving car. We just don’t know. But if they keep pushing the envelope with X, they might just get their outsized grand slam return that makes $1B of quarterly R&D “losses” look like walking around money.
Let that be a lesson to all startups and entrepreneurs out there. The game never changes. Only the stakes. Just keep going to work every day and let the power of compounding produce a benefit.
II. Surveying the Moon(shots)
Have you ever been to the Google[X] website? If not, it gives an overview of prior projects that they’ve killed, current in-flight projects, and projects that they’ve “graduated” to independent operating companies. Below is a summary.
Killed Projects. You learn more from the things you fail at than the ones you win.
- Foghorn (3 years): the product is fuel made from seawater to reduce the impact of fossil fuels on our environment. It shut down because it cost more to produce the fuel than you could sell it for. In other words: unit economics.
- Calcifer (1 year): an air ship for places with really expensive and poor shipping routes. It worked, but proved too expensive to get to a production-ready product. They also couldn’t find a good enough use case that people were willing to pay for to offset the cost. In short: unit economics + market demand.
Graduated Projects. Less risky endevours that found enough traction to be spun out of R&D and into an operating business of its own.
- Verily: their smart contact lens is one product you may have heard about, but the main premise of the company is adding software to health (e.g., bio-molecular nanotechnology, surgical robotics, etc).
- Google Glass (now called Aura): after apparently poaching a number of engineers from Amazon’s innovation arm, Lab126, this product lives on for enterprise-grade applications and have likely extended it to be more in line with the current visor form factor. Think jet-engine machinery repair at places like GE or more consumer oriented approaches like my thoughts on Apple’s upcoming augmented reality product.
- GCam: improving mobile photography. Looks like some machine vision implications.
- Google Watch: now called Android Wear, it’s a direct competitor to the Apple Watch. We’ll have to wait until September to see how the hardware stacks up head-to-head. Google’s pushing cellular connectivity within the watch itself. On the other side my guess is Apple wireless ear pods + Apple music enabled by the cellular chip. Plus faster processors in both (including wearable GPUs).
- Project Insight: indoor mapping. At first glance this doesn’t appear to be a big problem, but I saw initial marketing documents on how Apple’s Watch was supposed to have altitude sensors in version 1 of its hardware. A wearable like that can track which floor you’re on in a building. Google has the biggest maps database in the world, Foursquare the biggest points of interest, but nothing for indoor spaces exists at scale. There are, however, a few players trying to map outdoor 3D skylines so drones don’t run into buildings.
- Google Brain: this is the artificial intelligence lab working on machine learning, now a major part of Google. Sundar has said that we’re going from a mobile-first world to an AI-first world. On a personal note, I almost went to work here about 3 years ago. Deep Learning is nothing more than a math equation. But it will reach a fundamental brick wall because it’s not biological. It won’t be able to handle real-time streaming inputs without a database or training data to model it. But it seems everyone in the AI community is missing it because Deep Learning is so new and has so many magical applications. We’ll get to this more in a future post.
- Flux: design buildings of the future with sustainability in mind. Now that it’s an independent project, it’s created an open database for aspiring CAD architects. 3D print a building? Maybe one day soon.
- Project Tango: this appears to me like a similar concept to Magic Leap’s. Basically, use heavy machine learning to add to what you see through the lens of your phone. Of course, this is only applied as you move your phone around. This is not a special set of visors, but expect that to be on the roadmap.
Projects Still in R&D. These are the most exciting projects because they can twist, turn, and get pushed to fail in epic fashion. They might still die. But then again, they might still live.
- Self-Driving Car: click that link and view the favicon in your web browser. That’s a delightful little touch. I linked to this in a previous post but pay special attention to their monthly report PDFs. The most recent includes stats on the miles their cars have driven, as shown below. Did you know Google is the only one of the major autonomous vehicle companies who has reached Level 3 (no need for a driver). Tesla is still Level 2 (driver needs hands on the wheel at all times). Comma.ai is getting there.
- Project Loon: developing nations don’t have much, if any, internet access. They can’t afford underground fiber nor cellular towers for connecting to the internet via feature phones. So Google created a balloon. They’ve been working on this for years, but I think Facebook’s AI drone with laser beam internet is a better approach to connecting the whole planet. Either way, these projects are win-win for humanity so I give them two robotics thumbs up.
- Makani: creates clean energy through a kite. Sounds silly, eh? Have you seen pictures of those giant wind farms with a bunch of white propellers twirling around and around? The concept here is the same. Only instead of spending all that money and using all that metal, you just fly a kite that catches the wind and spins something to generate electricity.
- Project Wing: autonomous aircraft. Just don’t call it a drone. Because it’s bigger than that. They’re starting with the delivery use case, but I imagine the real moonshot here is delivery of people, not just products. They’re trying to do it cheaply and sustainably.
III. What’s Next?
Google called it X because that’s what marks the spot. Big markets. Big payoffs. Big vision. You start to get a very real sense of the focus and direction of the people who work at Google based on the products they’re trying to build above. Most have to do with sustainability.
Whether it’s adding more information to the world around us, moving things around our world, or finding better ways to generate energy, the one area the moonshot factory hasn’t seemed to touch is one where the moon actually exists. Space.
In today’s tech world, where artificial intelligence, self driving cars, planes, and rocketships to Mars are no longer science fiction but real products, what does that mean for Google’s next X?
What’s a bigger vision than terraforming and inhabiting Mars using our sun as the fuel source and asteroid mining for metals and water? The only thing it seems we don’t have an answer for yet is teleportation and warp drives. But in both cases, all we’re really talking about is moving things from one place to another.
If we simplify all of future tech down to it’s most basic two principles, is it really transportation and energy? Objects and humans are nothing more than data encoded in molecular structures that can be turned into information. Information can be encoded as data, then as 1s and 0s, and mirrored across the universe using quantum entanglement. Energy can be collected from places it already exists in the solar system.
On the longest time horizon, if energy can neither be created nor destroyed, then the only thing left is Time Travel. Though, that might just tear a rip in the fabric of space time, but heck, Alphabet’s earnings calls would be more fun when you hear some analyst ask,
What sort of risk mitigation strategy do you have in place for space time ripples? And in that alternate universe you described, can we pick the one where Alphabet’s stock price increases?
from Sean Everett on Medium http://ift.tt/2av9mym