Revealed: Elon Musk's Plan to Build a Space Internet

Musk wants to wire the world—and one day, Mars—using satellites

Because he doesn't have enough going on, Elon Musk—he of Tesla Motors, SpaceX, SolarCity, and the Hyperloop—is launching another project. Musk wants to build a second Internet in space and one day use it to connect people on Mars to the Web.

Musk is tonight hosting a SpaceX event in Seattle, where the company is opening a new office. The talk will mostly be about SpaceX’s plans for hiring aerospace and software engineers in the Pacific Northwest to boost the company’s rocket-building efforts. But he'll also use the talk to announce his newest idea, which would launch a vast network of communication satellites to orbit earth. The network would do two things: speed up the general flow of data on the Internet and deliver high-speed, low-cost Internet services to the three billion-plus people who still have poor access to the Web. “Our focus is on creating a global communications system that would be larger than anything that has been talked about to date,” Musk told Bloomberg Businessweek ahead of the announcement.

The Space Internet venture, to which Musk hasn’t yet given a name, would be hugely ambitious. Hundreds of satellites would orbit about 750 miles above earth, much closer than traditional communications satellites in geosynchronous orbit at altitudes of up to 22,000 miles. The lower satellites would make for a speedier Internet service, with less distance for electromagnetic signals to travel. The lag in current satellite systems makes applications such as Skype, online gaming, and other cloud-based services tough to use. Musk’s service would, in theory, rival fiber optic cables on land while also making the Internet available to remote and poor regions that don’t have access.

“Our focus is on creating a global communications system that would be larger than anything that has been talked about to date.” 

In Musk’s vision, Internet data packets going from, say, Los Angeles to Johannesburg would no longer have to go through dozens of routers and terrestrial networks. Instead, the packets would go to space, bouncing from satellite to satellite until they reach the one nearest their destination, then return to an antenna on earth. “The speed of light is 40 percent faster in the vacuum of space than it is for fiber,” Musk says. “The long-term potential is to be the primary means of long-distance Internet traffic and to serve people in sparsely populated areas.”

This project, he says, will be based in the Seattle office. (Musk has yet to determine the location of the satellite factory.) The office will start with about 60 people and may grow to 1,000 within three to four years. The employees will also work on SpaceX’s Falcon rockets, Dragon capsules, and additional vehicles to carry various supplies (and soon, people) into space. “We want the best engineers that either live in Seattle or that want to move to the Seattle area and work on electronics, software, structures, and power systems,” Musk says. “We want top engineering talent of all kinds.”

Earlier this week, the entrepreneur Greg Wyler announced a similar effort through a startup called OneWeb. Wyler has spent the last 15 years trying to bring Internet access to the so-called “other three billion.” He started a telecommunications company in Rwanda that set up Africa’s first 3G cell network. Later, he founded a company called O3b, which owns a satellite network that delivers fast, cheap Internet to hard-to-reach places along the equator. Through OneWeb, Wyler looks to expand this vision and fill the skies with hundreds of satellites that will beam their signals down to low-cost, solar-powered rooftop antennas.

OneWeb has announced that Qualcomm and the Virgin Group will invest in its effort, which is expected to cost around $2 billion. Wyler has also already secured the spectrum needed to deliver such a service from space and expects to be up and running by 2018. He has a team of more than 30 engineers developing the satellites, antennas, and software for OneWeb.

Musk and Wyler have known each other for years. Musk, in fact, used to crash at Wyler’s guest house in Atherton, Calif. While there are major similarities between the two ventures, Musk says he’ll have an edge through SpaceX’s smarts and manufacturing techniques. “Greg and I have a fundamental disagreement about the architecture,” Musk says. “We want a satellite that is an order of magnitude more sophisticated than what Greg wants. I think there should be two competing systems.”

Musk describes his system as “a giant global Internet service provider” for anyone. But he wants to go even bigger than that: He sees it as the basis for a system that will stretch all the way to Mars, where he plans to set up a colony in the coming decades. “It will be important for Mars to have a global communications network as well,” he says. “I think this needs to be done, and I don’t see anyone else doing it.”

"We see it as a long-term revenue source for SpaceX to be able to fund a city on Mars."

The backers of OneWeb, including Virgin chief Richard Branson, contend that Musk doesn't have the rights to spectrum he’ll need to create such a network. “I don’t think Elon can do a competing thing,” Branson says. “Greg has the rights, and there isn’t space for another network—like there physically is not enough space. If Elon wants to get into this area, the logical thing for him would be to tie up with us, and if I were a betting man, I would say the chances of us working together rather than separately would be much higher.”

SpaceX will need to be careful with its ambitions: Satellite makers have a choice as to whose rockets carry their machines. According to Musk, SpaceX will focus on making satellites for itself for the time being, rather than competing with its customers, although that may change over time. “I think we would consider building satellites for ourselves and for other people,” he says. “We’ll start by building ones that address the specific application that we are working on, and then we will be more than happy to sell to other people.”

Musk said it will take many, many years to have his Internet service up and running. “People should not expect this to be active sooner than five years,” he said. And it’ll be expensive: Around $10 billion to build, he says. “But we see it as a long-term revenue source for SpaceX to be able to fund a city on Mars.”

Google Wants Kids to Search the Darndest Things

Kids say the darndest things — and what they do say may make Google's search engine something of a mind reader.

Search engines already finish your sentences. Start typing "president" into Google or Bing or Yahoo and you're immediately presented choices you may have been thinking of. But, as you see from the examples below, the choices vary widely:

They also assume some pretty non-obvious things, like I am interested in politics, or in a particular president — or, as in the case of Bing, that I am a Turkish national?

Also, that I am probably an adult.

What if Google knew more about me before I touched a key?

A newly-revealed initiative by Google may be an important building block for an enhanced search engine that could further zero in on what I really want to know. In a lengthy interview with USA Today, Google VP of Engineering Pavni Diwanji talks about a "Google for Kids" initiative that would tailor some products for the under-13 set.

While she talks about protecting and serving kids (and their parents) it's not hard to connect some dots. This is a powerful way to achieve something it failed to do with Google+ — learn who its users are.

Diwanji recounts a story involving her own daughter, who pointed out the biggest flaw in the world's most-used search engine. It happened when she Googled 'trains.' "She came to me and said, 'Mommy, you should tell Google about Thomas the Tank Engine, because Google obviously doesn't know about him.'"

The secret to robust search is context, but greatness in search is requiring the searcher to provide as little ad hoc context as possible. It's a good first guess that you might be searching for something others are — hence those predictive type ahead suggestions.

But with more implied context, search can serve up better results. And nobody wants to craft Boolean sentences to refine the yield. We all want to live by the motto: "Do what I mean — not what I say."

Children are a great focus group. Kids searching the darndest things could inform other customizations that will yield results based on the data about me and only me: that I am a fireman, don't eat meat, or own a classic car. Relevant things that I wouldn't even think to mention in any given search session.

Backchannel Editor in Chief Steven Levy, author of "Inside the Plex," sees this as a natural fit. "What Google seems to want to do for children is very much in the spirit of what it wants to do for all of us — show us personalized results," he wrote in reply to an email.

Sure, there are privacy concerns (which Diwanji acknowledges), especially when it comes to targeting children. Google's entire business model is about looking over our shoulders to profile us for marketers, which is why special care and special laws are required when children are involved.

But I think this is a right-headed effort that doesn't exploit children. It leverages their unique thought processes to come up with answers for questions we wouldn't know to ask without them. The array of Google products tailored for children (not clear yet what they will all be) will be controlled by a parental dashboard, not by little Billy and Mary.

Is this right-track / wrong track? Will — should — the other search engines do something similar?

Enabling The Next Internet Wave In China

The leaders of China’s Internet success stories are rapidly becoming part of China’s establishment, interacting frequently with senior government leaders. Last week, for instance, several China Internet CEOs were part of Xi Jinping’s trip to Seoul. As they have these opportunities to spend time with policymakers, what should they be arguing is necessary for the next era of China’s Internet economy to be as successful as the one they have already profited from so well?

• Enhancing workforce skills. It should be obvious to all that universities should be graduating students equipped with the technical and business skills to succeed in operating companies. It is, however, entirely obvious today that they are not. Government action to make university curricula relevant for the growth in internet related jobs is.
• Sustaining openness to international ideas and capital. China has profited greatly from an Internet industry structure that encouraged the development of China relevant business models and local champions to deliver them. Yet often inspiration and stimulation for the business model came from outside China. And today, much of the world is looking to China’s internet leaders for their own inspiration. There is very much a two-way flow of ideas and investments. This benefits everyone.
• Enhancing data protection for individuals from businesses who are using their data. Many private companies and state-owned enterprises hold vast amounts of information on their customers. While new laws have been passed regarding how they are allowed to use such data, what permissions they should obtain, and the like, consumer confidence in the system is low. It could take only a few high profile breaches of personal information (e.g. from a bank or a telco) to have Chinese citizens pull back from providing their personal data so freely online. It is essential that China’s leading Internet companies are seen to be role models in data protection.
• Encouraging national markets. Most Internet businesses are born national. Yet regulations at a city or provincial level can hold back the development of efficient national markets that would benefit consumers. Constraints on selling second hand cars across provincial boundaries is one example. Policymakers should roll back regulations that constrain markets to the provincial or city level.
• IP protection. Increasingly, China’s internet leaders are developing a substantial amount of in-house intellectual property. They want to be certain they can protect this IP in China and internationally as they globalize. Ensuring they can get swift redress when they find their IP being used by others is key. For an Internet player guilty of using someone else's IP, a small financial fine in 12 months is almost irrelevant when a business is growing at Internet speed. Courts need to decide quickly if a business model is legitimate or not.

Swift action on these levers will place China’s Internet industries in a much stronger position to succeed going forward.

Your Right To Be "Un-Googled" And The Irony Of The Case

There is more than a touch of irony about Mario Costeja Gonzalez’s victory over search giant Google in the European courts, which grants him the “right to be forgotten” in Google search results.

Having successfully fought to have links to news reports of his financial affairs removed from search results, on privacy grounds, his name is unlikely to ever be forgotten now.

This week the European Union Court of Justice found in Costeja Gonzalez’s favor when it ruled that links to irrelevant or out-of-date information about individuals should be removed, on request. He had argued that issues raised in two articles in the newspaper La Vanguardia – and indexed by Google – mentioning that his house had been repossessed due to his social security debts, were out of date and no longer relevant.

His case is timely as support is strong in Europe for a “right to be forgotten”, giving anyone the right to have their personal information removed when it is “inadequate, irrelevant or no longer relevant”. The court-ruling, which can't be appealed, appears to show that European law supports this right – and Google (as well as a lot of other big data hoarders) could soon find themselves faced with a flood of similar requests.

Apparently, the Spaniard is not worried that his name will now feature far more prominently than it previously did in Google results for many years to come – in fact he has been clear that he is happy to have his name associated with the case. As he told the Guardian newspaper: “I was fighting for the elimination of data that adversely affects people’s honour, dignity and exposes their private lives. Everything that undermines human beings, that’s not freedom of expression.”

Google of course sees things differently – having called the decision “disappointing” – and argued during proceedings that such actions could amount to censorship. The law will not only apply to Google, but to all other search engines including Yahoo and Microsoft's Bing.

The “right to be forgotten” proposals in Europe in many ways reflect the recently passed Californian “eraser” law – requiring tech companies to remove material posted by a minor, if they request it – due to take effect next year.

The landmark case may be settled but of course debate will rage on over the issues at its core for some time yet. And it is unclear at this stage how it might apply to social media companies such as Facebook or Twitter. Does this mean that anyone who doesn’t like something which has been said about them online can demand that it is struck from the record?

At the moment, this seems unlikely – the court was clear in its ruling that publishers will have various defences – including public interest – with which to resist requests for information removal.

I hope therefore that it means criminals or corrupt public officials won’t be able to simply ask for links to news articles about their crimes be deleted simply because they would rather people forgot them. My fear is that there is a grey area in all of this. Some might see it as legitimate that a restaurant ask for bad reviews to be taken down if they are “no longer relevant” following a change of chef? Or a builder ask for his terrible feedback on the local “trusted trader” websites to be consigned to oblivion? The danger is that is could open the door for rogue traders to clean their sheet.

Campaign group Index on Censorship has been forthright with its criticism of the decision – calling it: “Akin to marching into a library and forcing it to pulp books.” For me, it all depends on how this new law (which currently only applies to citizens of the European Union) is implemented in practice. It will surely make life more complicated and expensive for search engine providers.

Whichever side of the fence you’re on, this ruling could have a big impact on anyone involved with storing or publishing data, and I’d love to hear your views. Do you see this as an important step toward more control of your public data or do you see it as censorship? Please share your views...