Day: October 16, 2018

Yang Kaili, a young internet star in China, was detained for five days for singing the country’s national anthem in a non-serious manner during a livestream. On October 7, clad in a fuzzy antler headband, Yang waved her hands around as she sang the national anthem for about 10 seconds.

“The national anthem is an embodiment and symbol of our country, and all citizens and organizations should respect and defend the honor of the anthem,” Shanghai police said in a social media announcement, the Washington Post reported. “Live-streaming webcast is not lawless territory and users should obey the law and uphold moral standards. The police will resolutely crack down on such behaviors that challenge the legal bottom line or public order and good social morals, in order to purify the Internet’s public sphere.”

Yang broadcast the short rendition of “March of the Volunteers” on Huya, a massive livestreaming platform. Yang’s internet celebrity first took hold on TikTok where she reportedly had over 44 million followers. She then ditched the lipsyncing app for Huya, where she also reportedly had over 2 million followers. Her Huya channel was banned following the national anthem incident.

“I sincerely apologize for the fact that I did not sing the national anthem in the live broadcast in a serious manner,” Yang wrote on her Weibo account, the New York Times reported. “My behavior deeply hurt everyone’s feelings. Sorry. Sorry to the motherland, sorry to my fans, sorry to everyone online, sorry to the platform.”

She also reportedly declared that she’d stop livestreaming work in order to focus on improving herself, studying laws related to the one she violated, and even watching national propaganda.

China enacted and amended the China’s National Anthem Law last year, which allows for imprisonment for up to three years for “deliberately distorting the lyrics or music of the national anthem of the People’s Republic of China, singing the national anthem in a distorted or derogatory fashion, or insulting the national anthem in other ways.”

The law doesn’t explicitly state which actions (or headwear) violate the law, but Yang’s detainment and the subsequent statement from police indicate that both internet celebrities and livestreaming platforms are not immune to being policed for seemingly wholesome, albeit silly, content. It also signals the government’s push for steadfast patriotic allegiance—or else.

[The New York Times]

California’s net neutrality law will cause “significant lost revenues” for Comcast, the nation’s largest cable company said in a court filing this month.

Comcast described the net neutrality law’s potential impact on its ability to charge online service providers and network operators for network interconnection.

“The paid interconnection provisions will harm Comcast’s ability to enter into new, mutually beneficial interconnection agreements with edge providers that involve consideration, leading to a loss of existing and prospective interconnection partners and significant lost revenues,” Comcast Senior VP Ken Klaer wrote in the filing in US District Court for the Eastern District of California. (“Edge provider” is the industry term for websites and other online platforms, such as Netflix and Google.)

Comcast submitted its filing on October 3 as part of the broadband industry lawsuit that seeks to overturn California’s net neutrality law (SB 822), which is slated to take effect on January 1, 2019 unless the court grants a stay halting implementation. Comcast’s filing is meant to support the industry’s request for an injunction that would halt enforcement of the law while litigation is pending.

California enacted its net neutrality law because the FCC repealed federal net neutrality rules at the request of ISPs including Comcast.

Not quite a ban on paid interconnection

The California law’s actual impact on network interconnection deals is not clear-cut. The law explicitly prohibits blocking and throttling and bans paid prioritization in the “last mile” portions of broadband networks that lead directly to consumers’ homes. The law also prohibits ISPs from requiring payment in exchange for delivering Internet traffic to consumers.

But the law applies a different standard to network interconnection, a type of traffic exchange that happens at the edge of each broadband provider’s network. The law says that ISPs like Comcast may not engage in interconnection practices “that have the purpose or effect of evading the prohibitions” of the net neutrality law. In other words, Comcast can’t block or slow down traffic at interconnection points, and it can’t make paid interconnection the only viable option for reaching Comcast customers.

Websites can reach Comcast customers without paying for interconnection, but buying the interconnection service ensures a smoother path into the Comcast network.

Comcast famously courted controversy when it required Netflix to pay for interconnection. Netflix has been paying Comcast since 2014, but previously argued that interconnection should be free since Comcast customers were requesting Netflix traffic while Netflix bore the cost of building a network to deliver that traffic directly into Comcast’s network.

Although the California law would not ban payments such as what Netflix pays Comcast, Comcast argues that the law is ambiguous enough “that some edge providers and even intermediary networks, such as CDNs and transit providers, will assert that the law prohibits Comcast from receiving consideration for interconnecting and exchanging Internet traffic.”

These intermediary networks “will contend that Comcast and other ISPs must provide unlimited free capacity to them for their delivery of other parties’ (i.e., edge providers’) traffic onto Comcast’s networks,” Comcast argued.

The law will make it difficult for Comcast to negotiate new interconnection agreements with edge providers, Comcast wrote:

Comcast has been engaged in discussions with at least two large edge providers about entering into new interconnection agreements that would involve monetary consideration. The prospect of SB 822’s enactment and its ambiguous “prohibition” on paid interconnection have already delayed these commercial negotiations. If the law becomes effective, it will likely be the death-knell of these commercial discussions, as it would put Comcast in the untenable position of risking an enforcement action or litigation by entering into interconnection agreements that could be claimed—and ultimately deemed by the California Attorney General or a California state court—to violate SB 822’s paid interconnection provisions. Our ability to negotiate new paid interconnection agreements with other edge providers will likewise be harmed.

The law could put Comcast’s existing financial agreements with edge providers “under a similar legal cloud,” Comcast said. The California attorney general or a company that pays Comcast for interconnection “could contend that the compensation provisions in our existing interconnection agreements are void or unenforceable,” Comcast wrote.

“Edge providers could seek to legally compel us to provision additional network capacity without any reciprocal consideration, or refuse to pay invoices for interconnection and force us to sue for non-payment,” Comcast also wrote.

Comcast claimed the law “will harm our customers, resulting in a loss of good will and damage to our reputation as an ISP.” It’s not clear how much revenue Comcast would lose, as the company said, “The loss in revenues and other monetary damages that the law will cause to Comcast will also be significant and are difficult to calculate.”

Comcast said it has about 100 interconnection customers, “including with many large edge providers and with CDNs such as Akamai, Limelight, and Cloudflare.”

Comcast reported $21.7 billion of revenue and $3.2 billion of net income in the most recent quarter. It’s not clear how much of that comes from interconnection, but it’s likely a small percentage—most Comcast revenue comes from service fees paid by cable customers and its NBCUniversal subsidiary.

AT&T also objects

AT&T also complained about the California law’s interconnection provisions in a court filing, saying that the law “purports to regulate such [interconnection] agreements and may be construed to prohibit payment for direct interconnection in many circumstances.”

“Enforcement of this new legislation threatens to involve regulators in the minute details of carefully reticulated interconnection agreements—deciding, for example, when particular networks are entitled to ‘free interconnection,’ at what physical locations, and at what levels of capacity,” AT&T wrote. “Such unprecedented regulatory intervention would threaten the interconnection arrangements supporting the modern Internet and impair network performance to the detriment of network operators and their customers. That outcome would not only disserve the public interest, but cause AT&T substantial irreparable harm.”

AT&T noted that it has interconnection agreements “with providers such as Facebook, Netflix, Google, Amazon, and Apple” and said the law could “cast doubt on whether AT&T should continue providing interconnection services under these contracts, which presuppose that AT&T will be compensated for the costs it incurs.”

Comcast vs. Cogent

Comcast spent much of its filing discussing Cogent, a network operator that exchanges traffic with Comcast in about a dozen US data centers and has refused Comcast’s payment demands. Comcast said it fears that Cogent will “claim that the ban on paid interconnection extends to transit providers like Cogent directly or as potential agents of or ‘proxies’ for edge providers.”

Cogent operates a global network and has direct connections with many major ISPs. Edge providers pay Cogent for transit, and in turn Cogent distributes their traffic to wherever it needs to go.

Internet traffic flows in both directions between Cogent and Comcast. Comcast argues that Cogent should have to pay Comcast because Cogent sends more traffic into Comcast’s network than vice versa, and Cogent argues that Comcast should have to pay because Comcast doesn’t operate a global network. The result is a stalemate with the companies exchanging traffic without payment, but Cogent occasionally complains that Comcast isn’t upgrading its side of the network connections quickly enough.

Comcast’s court filing said the California net neutrality law will benefit Cogent:

Notably, reading SB 822 to ban paid interconnection, as some entities will no doubt do, is unlikely to save most edge providers any money. Rather than connecting directly to Comcast, they will now likely pay a third-party transit provider, such as Cogent, to reach Comcast’s network. So while middlemen such as Cogent will make money by routing additional traffic onto Comcast’s network, Comcast and other ISPs will not be able to offer these edge providers competing paid interconnection services, which could be more efficiently and economically desirable for both parties.

Comcast said it would thus have to pay more for network upgrades, “which will shift the costs onto Comcast’s end users.”

Cogent responds

Cogent CEO Dave Schaeffer was, as always, quick to challenge Comcast’s assertions. Cogent generally only exchanges traffic for free with other global network operators that can also distribute traffic to the entire Internet, he said. Since Comcast only operates a US network, it should have to pay for transit to reach the rest of the Internet, he said.

“They do not meet Cogent’s requirements [for free peering] yet we do peer with them, and it’s only because of their ability to hold their customers hostage and use that market power to gain connectivity,” Schaeffer told Ars.

Comcast’s court filing is “totally silent on how they get connectivity to the rest of the world,” Schaeffer continued. “They claim they built out a US network but make no claim of an international network. They state they do not buy transit. What they in fact do is force other networks such as ourselves to carry that traffic for them for free. Their allegation of us getting a free ride is really in fact what they are doing, as opposed to bearing the cost of operating a global network.”

Comcast argued in its court filing that paying Comcast for interconnection can ensure better performance for edge providers such as Netflix “due in part to fewer hops over which the traffic must travel.”

Schaeffer said there are disadvantages for edge providers that purchase interconnection from Comcast instead of buying transit from Cogent or other transit providers.

“You have to pay Comcast a higher price per megabit,” and interconnection with Comcast provides a path only to Comcast’s customers, he said. “You still have to buy transit to get to the rest of the Internet.”

Some edge providers purchase interconnection from Comcast because “they’re afraid without that they have no way of reaching Comcast customers” when the other paths into Comcast’s network are congested, Schaeffer said.

Without regulation such as California’s, Comcast can delay upgrades in order to let its connections to Cogent become congested, Schaeffer said. To avoid congestion when bandwidth needs increase, both Comcast and Cogent have to upgrade the ports that allow traffic to pass between them.

“They reluctantly honor their agreement with Cogent and they do upgrade capacity, but they do it in a very slow manner and it usually results in some level of congestion,” Schaeffer said.

But after California passed its net neutrality law, Comcast “began to turn up those additional ports much more quickly,” he said. Comcast accelerated the upgrades nationwide, not just in California, he said.

“I think the importance of California as a state has forced them to modify their performance everywhere,” he said.

Posted by

Nathan Kirsch |

Tue, Oct 16, 2018 – 9:42 AM

NVIDIA has touted how deep learning technology on their Turing GPU architecture will change gaming and is finally showing off some detailed benchmark results. NVIDIA is using the Infiltrator benchmark to show TAA (Temporal Anti-Aliasing) versus DLSS (Deep Learning Super Sampling) on Pascal versus Turing powered cards. The benchmarks were done on Windows v1803 at 4K with custom settings. 

Turing is able to use the GPU’s Tensor cores with DLSS and that frees up the CUDA cores to do other work. The performance results are pretty impressive as the GeForce GTX 1080 was getting 32 FPS with TAA and the Radeon RTX 2080 was getting 60 FPS. There is a massive performance difference between TAA on Pascal and DLSS on Turing, but the slide doesn’t show TAA performance on Turing cards for some reason. Developers have announced 25 game titles will support DLSS in the months ahead and we can’t wait for the patches to start coming out!

More about DLSS can be found on this NVIDIA blog post.

Amazon updated its best value Kindle with some of the coveted features found in its high-end, $249 Kindle Oasis. The new Kindle Paperwhite announced today has a thinner, lighter design that’s now waterproof, making it the first Kindle other than the Oasis to have an IPX8 rating.

Amazon last updated the Kindle Paperwhite in 2015, giving it a better screen without raising its price. Now, the newest Paperwhite appears to be a mix of the old model and the now-defunct Kindle Voyage (the latter disappeared from Amazon’s site about a month ago). It has a 6-inch, 300ppi touch display with five backlighting LEDs, and the new screen is now flush with the black bezels around it. It’s still a black slab, but now it’s just 8.18mm thick and weighs just 6.4 ounces.

The waterproofing means users can read on the beach, in the tub, and around water in general and not worry about the safety of their Kindle. Also like the Oasis, the new Paperwhite has Audible integration so users can pair a Bluetooth headset and listen to audiobooks from the device itself.

This is a convenient feature for avid readers and audiobook listeners, and it will also likely help Amazon’s Whispersync business. When a user buys either the Kindle or Audible version of a book, Whispersync lets them buy the accompanying e-book or audiobook at a discounted price. Those who like to go back and forth between the e-book and audiobook version of a story may end up using Whispersync more now that Audible integration is available on a more affordable Kindle.

But the new Paperwhite isn’t exactly like the Oasis—the latter has a 7-inch display and a different design, as well as an adaptive light sensor, page-turn buttons, and more color options. Like all Kindles, the new Paperwhite has a battery capable of going weeks at a time between charges.

Amazon also announced some software changes that will be available on the new Paperwhite. The company updated the Kindle homepage to include easier ways to find books you may want to read based on your reading history, as well as stats about your reading habits. Users will also be able to make customized reading settings like font, boldness, and orientation so they can easily switch between reading profiles when necessary.

Amazon dominates the e-book and e-reader market, but it does have new competition in the US from Kobo and Walmart. Incorporating some premium features into the new Paperwhite, Amazon’s best-selling Kindle device, only makes it better able to weather the competition from Kobo e-readers like the Clara HD.

The new Kindle Paperwhite is available for preorder today starting at $129 for an 8GB model. Amazon also offers a $159, 32GB model and a $249, 32GB model with included cellular connectivity. The 8GB models will ship starting November 7, with the higher storage models shipping sometime in the coming months.

Listing image by Amazon

It’s July 17 and the temperature in the West Texas desert is marching, predictably, toward 100 degrees. But the air is cool inside the one-story prefab building where Jeff Bezos, wearing a North Face hiking shirt and a cap emblazoned with an Amazon Robotics logo, is attentive, back straight, listening. It’s only Tuesday, but the week has already been eventful. His company’s annual Prime Day sale proved so popular (good) that it temporarily took down the website (terrible). Amazon workers in Europe are striking. And the previous day, the price of Amazon stock had hit a threshold that put Bezos’ wealth at $150 billion, making him officially the richest person since people started keeping score.

It’s a number so huge that the Amazon CEO can painlessly siphon off a billion dollars every year to fund his boyhood dream: his other company, Blue Origin. Bezos’ money, earned from Amazon, has paid for the building where he sits, the air-conditioning, and the 60-foot rocket lying on its side in a nearby hangar, waiting to be tugged to a launchpad and shot into the thermosphere. Also, the salaries of about 1,500 Blue Origin workers, including the 35 or so engineers in the room and another 10 or so on a video screen, dialed in from the company’s headquarters in Kent, Washington. As they run down the checklist for the next day’s launch of that rocket, the New Shepard, Bezos sits near the back, not checking his phone even once. He asks one question—do the helicopters that will track the rocket’s flight know that weather balloons will be in the area? (Yep. Check.)

Starting next year, Bezos plans to use New Shepard to send passengers on jaunts into space. Clad in cool Star Trek–style jumpsuits, customers will settle into a comfy capsule and shoot up over the atmosphere for a quick peek at their home planet through panoramic windows and a few moments of weightless ecstasy. Though Blue Origin hasn’t announced the fee, it’s been reported to be a couple hundred thousand dollars per head, and Bezos anticipates ramping up quickly to a few flights a week. But suborbital tourism is just the beginning of his vision for Blue Origin. The second part of his plan is already under construction in a giant factory in Cape Canaveral, Florida: an imposing rocket meant for orbit and beyond.

Bezos tends toward discretion when it comes to his businesses, but earlier this year he offered to usher me into Blue Origin’s sanctums, with one stipulation: I had to promise that, before I interviewed him about his long-term plans, I would watch a newly unearthed 1975 PBS program.

So one afternoon, I opened my laptop and clicked on the link Bezos had sent me. Suddenly I was thrust back into the predigital world, where viewers had more fingers than channels and remote shopping hadn’t advanced past the Sears catalog. In lo-res monochrome, a host in suit and tie interviews the writer Isaac Asimov and physicist Gerard O’Neill, wearing a cool, wide-lapelled blazer and white turtleneck. To the amusement of the host, O’Neill describes a future where some 90 percent of humans live in space stations in distant orbits of the blue planet. For most of us, Earth would be our homeland but not our home. We’d use it for R&R, visiting it as we would a national park. Then we’d return to the cosmos, where humanity would be thriving like never before. Asimov, agreeing entirely, called resistance to the concept “planetary chauvinism.”

That vision captivated a generation of space nerds, including Bezos, who believed it back then, as a brainy schoolkid. And he believes it now, with “increasing conviction” every passing year. Earth is destined to run out of resources, he explains patiently to anyone questioning his priorities. Humans need a plan B. While he readily concedes that building a space company qualifies as a cool adventure, the ultimate point, he always insists, is getting people to live in space. He often remarks with astonishment and disgust that there have never been more than 13 humans in space at one time. He’s out to change that, by creating the backbone needed for O’Neill’s millions, billions, maybe even a trillion people to reside off-planet.

He’s not the only tech magnate with his head in the stars. Though Bezos has touched many more lives than Elon Musk (lots more Prime deliveries than Teslas), Blue Origin has received far less attention than Musk’s private rocket company, SpaceX. (Twitter follower metric: SpaceX, 7 million; Blue Origin, 123,000.) That’s in part due to Musk’s personality but also to his rocket company’s longer list of feats. SpaceX has had 60 successful launches of its Falcon 9 rockets and employs 6,000 people. Blue Origin has proceeded more slowly and with less oomph. Other gazillionaires—Richard Branson and Paul Allen—are also funding startup space ventures. All four men talk of creating the basic infrastructure for easy access to space, kind of like the railroad or the internet.

Step one for Bezos, of course, is proving that his rocket won’t kill its passengers, and tomorrow’s flight, the company’s ninth, will test whether New Shepard can handle a suborbital emergency at the lip of space. As the flight readiness review wraps up, Bezos stands to speak. He doesn’t usually deliver prelaunch pep talks, but today he makes an exception.

“This flight,” he says, in a tone more college professor than football coach, “is getting us so close to getting humans in space. I just want to ask you guys, while you’re doing all this tomorrow, to take a moment and not forget to reflect on the meaning and the beauty of what you’re doing. We’re getting very close.”

So close—to achieving a feat that humans first pulled off in 1961. But doing so then took an all-out national effort, and the US government’s interest has dropped considerably in recent years. Bezos believes that safely delivering non-­astronauts into space can move us closer to realizing dreams that have moldered for decades, like moon bases and orbiting habitats. All setting the stage for an epic migration that won’t begin until well after Bezos, and the rest of us, are long dead.

Jeff Bezos remembers being 5 years old and watching the Apollo 11 moon landing on a black-and-white television. The event triggered a lifelong obsession. He spent his boyhood in Houston and moved to Florida by high school, but he passed his summers on his grandparents’ farm in rural Cotulla, Texas. There, his grandfather—a former top Defense Department official—introduced him to the extensive collection of science fiction at the town library. He devoured the books, gravitating especially to Robert Heinlein and other classic writers who explored the cosmos in their tales.

When he was a junior at Miami’s Palmetto Senior High School, his physics teacher, Deana Ruel, tasked the students with designing a piece of playground equipment. Bezos’ idea was to build one in low gravity. “One day I’m going to be the first one to have an amusement park on the moon,” he told Ruel. He promised her a ticket. For a newspaper profile, Bezos spouted O’Neillian talking points to a local reporter curious about his space obsession: “The Earth is finite, and if the world economy and population is to keep expanding, space is the only way to go.”

Bezos went to Princeton, where he attended seminars led by O’Neill and became president of the campus chapter of Students for the Exploration and Development of Space. At one meeting, Bezos was regaling attendees with visions of hollowing out asteroids and transforming them into space arks when a woman leapt to her feet. “How dare you rape the universe!” she said, and stormed out. “There was a pause, and Jeff didn’t make a public comment,” says Kevin Polk, another member of the club. “But after things broke up, Jeff said, ‘Did she really defend the inalienable rights of barren rocks?’ ”

After Princeton, Bezos put his energies toward finance, working at a hedge fund. He left it to move to Seattle and start Amazon. Not long after, he was seated at a dinner party with science fiction writer Neal Stephenson. Their conversation quickly left the bounds of Earth. “There’s sort of a matching game that goes on where you climb a ladder, figuring out the level of someone’s fanaticism about space by how many details they know,” Stephenson says. “He was incredibly high on that ladder.” The two began spending weekend afternoons shooting off model rockets.

In 1999, Stephenson and Bezos went to see the movie October Sky, about a boy obsessed with rocketry, and stopped for coffee afterward. Bezos said he’d been thinking for a long time about starting a space company. “Why not start it today?” Stephenson asked. The next year, Bezos incorporated a company called Blue Operations LLC. Stephenson secured space in a former envelope factory in a funky industrial area in south Seattle. Other early members of the team included Pablos Holman, a self-described computer hacker, and serial inventor Danny Hillis, who had crafted a proposal to build a giant mechanical clock that would run for 10,000 years. Bezos also recruited Amazon’s general counsel, Alan Caplan, a fellow space nerd. (“We both agreed we’d like to retire on Mars,” Caplan says.) These people were more thinkers than rocketeers, but at Blue Origin’s start the point was to brainstorm: Had any ideas been overlooked that could shake up space travel the way the internet had upended terrestrial commerce?

Another early participant was George Dyson, a science historian and son of physicist Freeman Dyson. At the 1999 PC Forum, an elite tech event run by Dyson’s sister, Esther, Bezos made a beeline for George, who had been writing about a little-known 1950s venture called Project Orion. Project Orion sought to propel space vehicles with atomic bomb explosions, and Bezos wanted to know all about it. As Dyson recalls, Bezos saw Orion as “his model for a small group of crazy people deciding to go into space without the restrictions of being an official government project.” (Bezos later reviewed Dyson’s book on Amazon—something he’s done only three times in the company’s history.) Some months later, Stephenson asked Dyson if he would consult for the company. Then he asked him to join Blue.

When Dyson signed on, he says, Blue Origin felt like Wernher von Braun’s Society for Space Travel. Like that amateur group of dazzling scientists, Blue resembled a club more than a company. Its members were obsessed with finding an alternative to chemical combustion, which is a woefully inefficient way to propel rockets on interplanetary journeys. “We went through a long list of not-quite-crazy but way-out-there projects at the beginning,” Dyson says.

Those were hashed out at Blue Origin’s monthly Saturday all-hands meetings. The sessions began at 9 and lasted all day. Bezos rarely missed one. “It was almost incomprehensible how technically engaged Jeff was in every part of the discussion,” Dyson says. “It wasn’t like, ‘Oh, we’ll leave the hydrogen-flow control valve question to the hydrogen-flow control valve people.’ Whatever the question was, Jeff would have technical knowledge and be involved.”

But as the Blue Origin team experimented with eccentric ways to heave things upward, they began to realize there was a reason big tubes full of chemical fuel had persisted. Every new tack proved infeasible, because of cost, risk, or technical complexity. “You can work really hard and come up with what you think is a super original idea, and you always find out that some Russian guy tried it 15 years ago,” Stephenson says.

By 2003, Bezos had started hiring experienced aerospace engineers, including some who’d worked on a ­government-backed project called the DC-X. It was a prototype of a reusable rocket that would be able to land vertically. Bezos says he studied alternative means of propulsion “for three years with a small group of brilliant people and concluded with complete conviction that rockets are actually not just a good solution for getting off the surface of the earth, they’re a great solution. But they have to be reusable.”

The logic is obvious: No one wants to junk a vehicle after a single flight. But recovering rockets has proven tricky. After exploring different ways to land them—parachutes, air bags, retractable wings—Bezos settled on vertical landings, using legs. Now it’s a core principle, almost a commandment at Blue. “Rockets have legs,” Bezos says. “Rockets should land! This is how God meant rockets to be!”

Musk was coming to the same conclusion. Seeing the similarities between the two companies, one Blue Origin employee, Tomas Svitek, urged Bezos to collaborate with SpaceX. Bezos and Musk met for dinner in the fall of 2003, but nothing came of it. “He’s a good guy, we’re kindred spirits,” Bezos told Svitek afterward. “But we decided to do our own thing.” Bezos now describes the meeting as more of a social event, a convivial dinner with their spouses. It’s fair to call this the high point of their relationship.

As Blue Origin grew, Bezos began to see it as the infrastructure for future space entrepreneurs to build even more exciting things. “There is no way two kids in a dorm room can build a super interesting and important company in space,” he says, because it costs way too much to get started. “If I can unleash a thousand Zuckerbergs in the next generation, we will see things you can’t even imagine.”

And what does that infrastructure look like? “Reusability, reusability, reusability,” he says.

“Rockets have legs,” Bezos says. “Rockets should land! This is how God meant rockets to be!”

For Blue Origin, that meant first learning how to land an object vertically. It didn’t have a rocket, so in 2003 it started building a stand-in, an awkward 20-foot-tall platform powered by surplus Rolls Royce Viper jet engines from the ’60s. Named Charon, after the ferryman in Greek mythology who transits deceased souls to Hades, it resembled a mini offshore rig, with cables snaking around dense scaffolding and a jet engine on each corner. Its jointed legs were designed to bend upon touching down, like shock absorbers. Using a crane, the engineers repeatedly hoisted the platform and then dropped it, to see if its legs could withstand the fall.

On March 5, 2005, at Moses Lake, about 180 miles from Seattle, the whole company showed up, along with spouses and kids, to see Charon fly. With burgers simmering on a giant grill, they watched as it rose up like a quadcopter drone. It hovered at 316 feet and then, with its autonomous software controlling the thrusters, lowered itself back to the flight pad, to lusty cheers from the Blue crew. An ebullient Bezos, in a blue hard hat, popped the cork of a jeroboam of champagne and splashed it into everyone’s paper cup. “To the Charon team!” he shouted, accompanied by his trademark braying laugh.

“It was like a junior high school kid’s project,” Holman says. But it was the proof of concept for “everything that’s followed.”

In the years after the Charon moment, Blue Origin built and flew actual rockets, including a 2011 test flight that ended with a dramatic accidental explosion. When it began sending up the New Shepard, in April 2015, Blue successfully separated the capsule from its booster, which houses the propulsion system, and then reclaimed the capsule. Later that year, New Shepard landed a booster for the first time. (When SpaceX landed its first rocket, eight months later, Musk got peeved at Bezos’ “welcome to the club” tweet. Bezos insists it was sincere and not a jibe.)

By then, Blue Origin was looking more like a traditional aerospace company. It moved from the envelope factory to a modern facility in Kent, about 20 miles south of Seattle. Using custom rockets built in Kent, Blue Origin has conducted nine suborbital flights. Bezos likes to wander the floor of the factory—he spends Wednesdays there—asking himself, “What would surprise Wernher von Braun today?” One twist that would impress the late German rocket scientist, he suggests, might be Blue Origin’s fancy new hydrostatic bearings. Spaceflight has a way of placing enormous strain on traditional bearings, the mechanical components that reduce friction between two moving parts. The ones Blue Origin developed rotate on a thin film of high-pressure fluid. Other points of pride are a new alloy, which Blue has named cascadium, and a coating it’s calling rainerium.

By most people’s metrics, however, Blue Origin lags behind SpaceX, which has placed dozens of satellites into orbit and carried cargo to the International Space Station. Bezos counters that Blue Origin’s pace is not a bug but a defining feature. The company’s mascot is a tortoise (leaving unspoken who the hare is). Its motto, translated into Latin, is Gradatim ferociter—“Step by step, ferociously.”

Even as Blue Origin was settling on its method of getting people into space, Bezos was seeking a location from which to pitch them into the sky. He combed through census data for remote areas and eventually settled on a 300,000-acre plot north of the tiny town of Van Horn, Texas. It borders the Sierra Diablo mountain range, where the last battle between the Texas Rangers and the Apaches took place. An old ranch complex there has been turned into a Bezos family retreat. And inside one of his mountains, a team is building the 10,000-year clock that Danny Hillis envisioned.

When I arrive at the Blue Origin complex on July 17, the day before the New Shepard flight test, Bezos acts as my guide. He gives his pep talk to the Blue team—which will later be included in a short, slick video posted on the company’s Twitter feed—and we breeze through the engineering area and the mission control room. We head out the door to a different building, a tall barnlike structure that houses the main attraction: the New Shepard. Today, a light indicates that the building is marked Code Amber; extra precautions are in place, and we must shelve our phones before we enter. As social media smart alecks have tirelessly noted, New Shepard looks like a dildo. The design reflects Blue Origin’s focus on passengers. As with any rocket, the bulk is devoted to its propulsion system, but its dome is blunt to house a roomy capsule. To seat six space tourists comfortably in reclining chairs (and to let them cavort in weightlessness), the capsule needed to have a 12.5-foot diameter, making it wider than the booster. The designers used this mismatch to their advantage, topping New Shepard with a wide ring that has panels to help keep the rocket vertical upon descent.

I reach out to touch the rocket and Bezos quickly but gently admonishes me. We climb up the scaffolding, taking care not to brush against the giant tube, which is painted with Blue’s signature large feather. Visible through a capsule window is Mannequin Skywalker, a dummy that Blue has already sent to space twice. If all goes well, sometime around 3 am workers will load the rocket onto a trailer, hitch it to a truck, and drive Skywalker and his ride to the launchpad about 2 miles away.

Tomorrow’s flight will test whether the crew capsule can execute a high-altitude escape: If something goes wrong as the rocket leaves the atmosphere, can the capsule speed itself away from an about-to-explode booster? It’s a bare necessity for space travelers, who will expect, at a minimum, to survive their excursions.

Bezos says that Blue will carry humans into space in the first half of 2019. Then it will launch its suborbital tourism business, perhaps before the year is out. Since the flight itself will be automated and designed for comfort, Bezos guesses that his customers will need only a day of training. They will be astronauts in the way that people who sign up with the Universal Life Church to perform marriages are clerics. Bezos anticipates that they’d sign a waiver—an FAA requirement—that would read like a form you’d sign for a scuba diving excursion. The six black seats, resembling souped-up barber chairs, are arranged in a circle, each with a view through giant windows. (No middle seats!) Computer displays will show footage of the takeoff, the booster and capsule separating, and other events. When the capsule crosses into space, the passengers will get about four minutes to unbuckle and float around inside. After that suborbital idyll, a recorded voice will instruct them to return to their seats.

When Bezos tells me the bit about the automated voice, I’m taken aback. No flight attendants? I imagine a lot of panicky fumbling as a bunch of first-timers try to strap back in for a rapid 300,000-foot-plus descent.

“There’s no flight attendant,” he confirms. “There doesn’t need to be. The vehicle is automated. It’ll probably be even easier to get into your seat in zero g—there are lots of handholds.” Bezos could be missing the need for a psychological handhold, the comforting presence of a human expert as you tumble around. I later predict to some Blue engineers that the first flights will indeed have human guides; they say they wouldn’t bet against it.

The whole trip will last about 11 minutes, making it seem like a very expensive Disney ride. It almost makes you wonder why, when Blue’s long-term goals are so high-minded, it is pursuing a project so seemingly frivolous. Bezos’ response is that tourism is a source of revenue, but also something bigger—the best way to make space travel seem routine. If people you know (or at least have heard of) are popping into suborbital space, it will start to feel more normal and less risky. Commercial air travel also followed this path, with early passengers engaging in fervent prayer on the runway. Meanwhile, Blue will use the trips to perfect its rockets. “We wanted a mission that would fly a lot,” Bezos says. “Launching commercial satellites, you’re lucky if you do it a couple dozen times a year.” (Bezos himself hasn’t yet committed to a flight. “I’m going for sure,” he says. “I don’t have a time frame.”)

After viewing the New Shepard in its hangar, we head to the launchpad, about 2 miles away on a road hacked out of the brush. A few fuel tanks, a low tower, and some cameras flank the concrete pad. Bezos isn’t done showing me around, though. We leave the Blue Origin base and travel to the top of the 6,000-foot mountain that contains his 10,000-year clock. The gigantic timepiece sits in a deep shaft, encircled by a steep spiral staircase. Descending into the cavelike quiet and viewing the exquisite metal gears in that quasi-monastic setting is almost a holy experience. As per Hillis’ design, the clock will emit a tick each year, each decade, each century, and each millennium.

Bezos is famous for taking a long view (Amazon became notorious for postponing profits), so it’s not surprising that he decided to fund the clock project, which will open for visitors in the next few years. “The clock is designed to be a symbol to encourage long-term thinking,” he says. “If people pay attention to the clock, they’ll do more things like Blue Origin.” As we exit through an underground passage to the mountain summit, we take in the stunning vista of his land, with the launch facilities of his rocket company discernible in the distance.

So Bezos’ domain spans both space and time. But to his critics, these lofty pursuits seem disregardful of all-too-pressing earthly concerns. They say the richest man in the world should be more invested in tackling climate change or extreme poverty or the diseases ravaging human lives—or just about anything else. There’s also a camp, with US senator Bernie Sanders as unofficial spokesperson, that thinks Bezos’ charity should begin in the workplace, not the cosmos. In a scathing fund-raising letter, Sanders said that for Bezos to bankroll a space company is “absolutely absurd.” When I ask him to clarify, the senator says, “I find it absurd that he has billions of dollars to spend on his space ventures but apparently does not have enough money to pay his workers at Amazon a livable wage right here on planet Earth.”

Bezos acknowledges that there are pressing needs. He says he’s a great admirer of Bill and Melinda Gates and how they have used their fortune to help humankind. He’s talked to them about the Giving Pledge, which commits its wealthy signatories to put more than half their riches to philanthropic causes, and says he is considering signing on. A few weeks after our conversation, he made his biggest contribution to earthly wellbeing by announcing the Bezos Day One Fund, which will devote $2 billion to helping homeless families and creating a network of Montessori-esque preschools. In early October, Amazon announced that it would begin paying all employees at least $15 an hour.

Bezos sees tourism as something bigger—the best way to make space travel seem routine.

But the way he sees it, he’ll make his biggest impact with his businesses. “I have the great luxury of my resources,” he says. “I’m not going to work on anything that I don’t think is improving civilization. I think The Washington Post does that, I think Amazon does that, and I think Blue Origin does that. In the long run, Blue is the most important.” Bezos often speaks of how easy it is to be misunderstood when working with extremely long time horizons. People will appreciate his crusade, he says, when the ravages of climate change, depleted resources, and unbreathable air make it time to discard what Asimov called, on that TV show Bezos enjoyed so much, “planetary chauvinism.”

Bezos’ infatuation with space habitats has been remarkably consistent since his high school days, even as O’Neill’s fan base has dwindled. In a book about the Princeton physicist, one writer described O’Neill’s vision as a “failed future.” Even some space habitat enthusiasts question whether now is the time to act on it. Sara Walker, an astrobiologist at Arizona State University, balks at the ecological engineering it will require. Since O’Neill’s time, we’ve become more aware of the mind-boggling complexity of life on Earth and our ignorance of how to re-create it in a giant floating can. “Is there a critical amount of biodiversity needed?” she says. “We don’t know.” Others posit ethical concerns, as in “Who the hell are we to impose our will on the universe?” That goes double when the “we” is rich white guys. (That Princeton woman incensed about the rape of the universe was ahead of her time.)

Bezos buys none of this. He sees humanity’s move to space as the only option. He sees it with the certainty of mathematics. So we might as well get started, long before we exceed the energy resources available to us and face catastrophe. “I’m just building the infrastructure so it’ll be there when people realize they need it,” he says. How will we know it’s time to blast off—will it be in one generation, or 10? “It will become incredibly obvious,” he says.

The solution, as Bezos sees it, is to get off the planet to better exploit solar power, so that the sun’s abundant photons can support the fruitful existence of countless people. (We’d also grow real fruit in space.) “Wouldn’t your grandchildren’s grandchildren’s lives be so much more exciting if there were a trillion humans in the solar system who used more of that output to do amazing things?”

“If you don’t agree with that,” he says, “forget the whole thing. AH-HAH-HAH-HAH.”

In Cape Canaveral, 1,400 miles from his Texas ranch, Bezos has built a new $250 million facility. No prefab here; it’s a statement building. BLUE ORIGIN is spelled out in huge block letters clearly visible to visitors at the nearby Kennedy Space Center. The lobby is a shrine to the company’s achievements: On the right is the rocket that has gone to space and landed on its feet five times. On the other side is a capsule with six tortoises painted between the windows, to mark each of its sorties.

It’s here that Blue Origin is building New Glenn. If New Shepard is Bezos’ ploy to get ­people comfortable with space travel, New Glenn is the company’s galactic workhorse. At 283 feet, it’s taller than SpaceX’s Falcon Heavy rocket and just shy of the Statue of Liberty. Its seven engines put out 3.85 million pounds of thrust, compared with New Shepard’s 110,000 pounds. New Glenn won’t make its first flight until 2020, but Bezos envisions a fleet of them performing a wide range of tasks—“commercial satellites, planetary missions, NASA missions, national security missions, everything.”

So far Blue Origin has signed deals with four companies, including Eutelsat, a French communications company that wants New Glenn to deliver a satellite. The company also plans to sell the engine that powers New Glenn to the United Launch Alliance, a joint venture between more traditional government contractors. New Glenn will wind up competing with SpaceX’s larger rockets, but so far there seems to be enough work for everybody. Though not enough primo launchpads, apparently—when SpaceX got a 20-year lease to use the historic 39A launchpad where Apollo 11 took off, Blue Origin unsuccessfully challenged NASA on the decision. Blue will launch from the less storied Launch Complex 36.

Bezos hopes that launch deals, government contracts, and, to a lesser degree, the fees from tourists will eventually make Blue profitable, though the time required may be best measured by the clock he’s building. “Right now, the business model with Blue Origin is I sell Amazon stock,” Bezos says. “I’m willing to be patient for decades.”

Meanwhile, the company is deep into the design of a lunar lander called Blue Moon, which is meant to haul 10,000 pounds of stuff (roughly equivalent to two Ford 150 pickup trucks), perhaps by 2023. “The moon is an essential place to get resources to build these kind of O’Neill-style colonies,” he says.

Those colonies again. I ask Bezos if he’d really want to live in one of those things.

“Yeah,” he says. “They’re not what you imagine. I mean, they’ll have farms and rivers and universities; they could have a million people in them. They’re cities. But I’d want to be able to come back and forth to Earth too.”

At nearly 10 am on July 18, New Shepard is about to launch in the Texas desert. I’m watching from a VIP area a few miles away, on the edge of a rocky canyon. My fellow gawkers include observers from the European Space Agency, plus-ones of Blue employees, and the Southwest Airlines pilot who safely landed a plane in Philadelphia after her engine blew up.

On the horizon, we see smoke pouring from a distant spot partially obscured by haze. A yellow-orange tail of fire pushes a needle into the sky, the low rumble of engines surfing through the plains and into the canyons. We crane our necks upward to follow the moving speck until all that’s left are vapor trails, blown by the wind into a pretzel shape. The rocket is on track for its 100-kilometer ascent. Though we can’t see it, we learn that the capsule has fired its motor for a quick escape from the booster, as planned. What remains to be seen is whether the maneuver compromised the booster’s return to its landing pad.

Minutes pass. Suddenly, out of the empty sky comes the thin needle of the descending rocket. It appears to be just plain falling. Because it remains vertical, it seems like an illusion, like a liftoff video played backward. Then suddenly, just as the rocket has decelerated to a hover, we see a bouquet of fire. A sonic boom smacks our ears, and I’m wondering whether a fiery explosion will make this Shepard flight its last. But the flames are simply the rocket slamming on the brakes, firing its engines to gently lower New Shepard onto its landing pad. A few seconds later, the capsule floats down on three parachutes, kicking up a small plume of dust as it lands.

A group of us pile into vehicles and drive to the headquarters. Bezos has gone ahead, leading a convoy to unite with the capsule and booster, which have landed only a quarter-mile apart. The nearest road ends a few hundred feet from the capsule, and we make our way on foot through the brush, with an eye out for rattlesnakes, to the cluster of Blue-sters already surrounding it. A crane is preparing to lift it by the nose for the ride back to the barn.

As the capsule comes off the ground, someone spots a remarkable thing: a living creature underneath the capsule. A horned lizard, frozen in confusion, has its mouth agape as if to say, “What the fuck?!” The reptile had been minding its own business, thinking whatever thoughts a sunbather might have on a hot July day, when an object from outer space landed on it. Bezos’ laugh booms from here to the Diablos.

---

Editor at large Steven Levy (@StevenLevy) wrote about Paul Allen’s space company, Stratolaunch, in issue 26.09.

This article appears in the November issue. Subscribe now.

Listen to this story, and other WIRED features, on the Audm app.

Let us know what you think about this article. Submit a letter to the editor at mail@wired.com.