Mastertronic, SIGTRAP Games, Sublevel Zero.

There's a robot that lives inside No Man's Sky that nobody outside of the development team may ever see, because its entire purpose is to fly to each of the game's 18 quintillion worlds, take short videos and document its interstellar travels as a series of animated GIFs.

Developer Hello Games built this probe because, at its core, No Man's Sky is so massive that its statistically insignificant team of four human artists can't control or oversee its creation. At best, they can audit some of what already happened and adjust accordingly.

So, instead of pure human oversight, No Man's Sky is governed by what art director Grant Duncan calls a team of thousands of "mindless idiots" — the computer code and algorithms that take the structure that he and his team created and turned that into a universe.

The decision to give up control wasn't easy. Duncan's now-loving embrace of procedurally generated art started as something that existed on the continuum between ambivalence and outright hostility.

But now, deep into the development of No Man's Sky, and despite his initial reservations, he learned to stop worrying and love procedural art, a process he documented in a GDC 2015 session this week.

The problem that probe is designed to solve is one that hits artists particularly hard: When every atom is procedural, developers must, by definition, lose control. Duncan learned to embrace this, at least in part out of the necessity to create a universe with planets numbering in the quintillions (a number so big that he admits to it being "entirely meaningless" to the human mind). Thus, the autonomous robot and the "idiotic" team of virtual art directors that keep tabs on the unfathomable number of celestial bodies, their flora, their fauna, their geography and their indigenous creatures.

The idiotic robots Duncan deployed are responsible for creating what players will see in No Man's Sky. This is no exaggeration. There was no other way to create the the open universe game set for release sometime this year on PlayStation 4 — and, presumably, later on other unspecified platforms.

It really is a universe, full of a myriad of stars. Each star has a solar system. And each solar system has planets. And planet-hopping — the process of traveling from planet to planet, solar system to solar system and galaxy to galaxy to explore and exploit — is perhaps the easiest gameplay mechanic to understand in No Man's Sky, a game whose whose other activities Hello Games has been deliberately tight-lipped.

Near the beginning, while creating the vast universe of No Man's Sky, Hello Games began with the idea of with planet-hopping. And when making something like that an intrinsic part of the game, they needed to do more than just create planets. They needed to fill them with plants and animals that made them feel alive.

To invent these alien worlds, the developers took inspiration first and foremost from something that might seem surprising: not games, not movies, not sci-fi novel covers, but the real world. Hello Games wanted No Man's Sky to be grounded in the "believability" of the real world, Duncan said, which is already filled with "bonkers" and "crazy" things, whether animal, vegetable or mineral. And if they wanted players to explore their creation (and they do), they had to make their universe and everything in it interesting and strangely familiar.

It may be a surprising twist, but it's not without logic. Just think of how many exotic and varied species of frogs exist in the real world, from the standard, green croaking lumps of fairy tales to the little neon-colored poisonous monsters of the rainforests. In No Man's Sky, reality established the rules. Then the artists went about breaking them to create alien creatures, planets and geographical formations that are, fundamentally, grounded in reality.

To populate this alternate reality, the artists created seeds: the essential parts of plants and animals and geographic locations. Trees have trunks and leaves. Animals have bone structures. Spaceships have cockpits. Buildings have doors, windows and roofs.

Then they threw those seeds into what Duncan calls a "big box of maths," where the British developer's algorithms create variations on those themes. Short trees with orange foliage. Spaceships with stubby cockpits. Alien creatures whose deer-like ancestry is graspable at a glance.

Feed No Man's Sky's big box of maths the same inputs — the art team's seeds — and it will produce the same outputs, whether on your system or a friend's. That's where its mighty universe comes from.

But that's the technical side of the equation. Duncan is an artist. It's his job to ensure that what the math box creates makes sense, looks pleasing and, though alien, feels familiar. And that happens at the intersection of art and math, a crossroad that makes some artists uncomfortable.

When he started working on procedural generation and No Man's Sky, he contacted several artist friends, many of whom shared the approximate opinion that, as Duncan put it, "procedural is a big pile of shit." It creates terrain so realistic that it's boring, they said. It creates blobby, cartoony creatures, they said. It takes control way from artists, they said. It's "programmer art," which isn't really art, they said.

Procedurally generated art requires the artists to surrender control

Moreover, if the Hello Games art team was just recreating things they found in the real world, that'd be boring. It's uninteresting to artists, who see procedural technology as random and, he said, "can never lead to good art or that produces "inferior visuals."

Duncan's job was, in a sense, to prove the doubters wrong by convincing them to let go.

"I think that the truth is, we're actually all control freaks," Duncan said. "Artists are so used to having complete control of every single pixel. Especially now with digital artists. We can get Photoshop, we can zoom right in and obsess over something no one will ever care about."

So the artists at Hello Games created a blueprint system, wherein the artist creates a template. It begins, in other words, with handmade art. Then the math box creates variations on those themes — it tweaks the lioness' leg length, stretches her neck, colors her.

The tiny team focused on efficiency, again out of necessity. To create as little work as possible, they identified similar things, like bone structures. The system can move bones around and create poses that range from owls to proud stallions and "hundreds" of different movement cycles. That applies to everything from ships to sea creatures. In No Man's Sky, dolphins, sharks and whales all share a common ancestor.

But Duncan's artist friends were right about at least one thing: feeling the box of math isn't enough. That process is mechanical, and math doesn't have the same artistic sensibilities that humans do. So the artists collaborated with the coders to make random generation more pleasing.

Now the box of math nows that, when it places a boulder on the ground, it should also place a few medium and smaller boulders around it. It makes sense. You get, he said, "a lot more pleasing silhouettes."

If you're going to create a universe for people to explore, you've got to make it believable.

That same system of symmetry applies to entire planets. One gets created, and it has certain properties. Those properties (say, a blue mineral) have an effect on the plants that grow on the planet and the creates that inhabit it, informing everything from their design to their complimentary colors. But before Hello Games' artists did that, they had to teach art to robots.

Finally, after the artists created the blueprints and planted the seeds, and after the coders constructed a box full of math, the universe exists. Someone has to look after it.

That's where the robot probe comes in, flying from planet to planet, making the animated GIFs, so that the folks at Hello Games can make sure everything looks acceptable. That there aren't too many red planets; if so, they can fly in and tweak things to ensure consistency and coherency.

Because if you're going to create a universe for people to explore, Duncan said, you've got to make it believable for players. And that, beyond just designing things, is the job of an artist.

"The best games to me" Duncan said, are those in which "you don't question anything. Things just feel like they belong together."

The rate of recidivism for sexual offenders is lower than that of most other crimes (pdf). But ever since the 2013 rape and murder of 8-year-old Cherish Periwinkle by a released sex offender in Jacksonville, Florida state legislation on sex offenses have become among the strictest in the nation.

Local sex offenders’ prison sentences can no longer be abridged for good behavior. Even after prison, offenders may be forced into indefinite periods of confinement called “civil commitment.” Tracey, 47, came to Miracle Village from a privately-operated civil commitment center in Arcadia, Florida.

While the number of sex offenders registered across the US has risen 23% over the past five years, sex crimes seem to be most aggressively identified and punished in Florida, where registrations have leapt 74%. More than 100 new sex offenders apply for a spot every day, says Miracle Village’s website. The community has renamed itself accordingly: in Jul. 2014, humble Miracle Village became the City of Refuge.

Miracle Village is a book by Sofia Valiente. We welcome your comments at ideas@qz.com.

Netflix already announced release dates for a slew of new shows today, but it's not stopping there: the company has acquired worldwide distribution rights to the newest film from the director of True Detective. Written and directed by Cary Fukunaga, the project tells the story of a young boy that's made to join a group of soldiers in West Africa. Netflix reportedly paid $12 million to snap up the film, which stars Idris Elba (Luther, Pacific Rim). According to Deadline, the film will be streamed worldwide on Netflix, but will also be released theatrically, with an eye towards promoting the film heavily for awards consideration.

It's just the latest move Reed Hastings' streaming giant has made into theatrical motion pictures over the past several months. Last year the company announced that it would be putting out a sequel to Crouching Tiger, Hidden Dragon as a combination Netflix and IMAX release, but exhibitors quickly protested at the notion of undercutting the traditional theatrical window with a simultaneous release. The furor didn't cause Netflix to miss a step, however. In January the company signed a deal with Mark and Jay Duplass for four new Netflix films that would also receive theatrical distribution; a month later it picked up a new film starring Fifty Shades of Grey star Jamie Dornan, and a week after that came word that the company would be producing a new sequel to Pee-wee's Big Adventure.

Netflix is playing the long game to wear down theater chains

Netflix appears to be very aware of the pushback it will face from theater chains, and is building out its theatrical strategy accordingly. The films it has stated it will release in theaters thus far are niche, awards-friendly films that will be extremely attractive to small, independent movie theatre chains — the same kind of theaters that haven't been dissuaded by day-and-date releases the way that that larger chains like AMC have been. And while Netflix is jumping in to produce broader, more mainstream fare like the Pee-wee sequel and several upcoming Adam Sandler movies, those films have been primarily earmarked as projects that will skip theaters altogether and go straight online. Should Netflix's original films end up being as consistently good as its series have been, it could put larger theater chains in an unfortunate bind come awards season: decide to play ball with the company's strategy, or miss out on the movie's everybody is talking about.

I'm looking for cherry flowers, like the beautiful pink ones blooming all over right now, but in a clean place, and not somewhere I'll get in trouble? I don't want to steal them, but I'd like to pick some to make mochi, jellies, teas, and tea cookies using the blossoms. I think it's not legal to take them from a park. Do you have any ideas?

Gone Home, a first-person exploration game that hit PCs back in 2013, was a breath of fresh air for gaming in that it demonstrated how to make an affecting game without ever needing a gun. Firewatch looks like its natural successor, as watching 17 minutes of gameplay, courtesy of IGN, is a fascinating and singularly engrossing introduction into the life of a fire lookout.

Firewatch is being developed by San Francisco-based out Campo Santo, which is comprised of folks who helped develop The Walking Dead Season 1 at TellTale Games, as well as minimalist artist Olly Moss, whose work you might recognize from a few certain Star Wars posters. The title follows Henry, voice by Mad Men's Rich Sommer, as he traverses his corner of the woods and winds up being pulled into strange occurrences. There are no guns.There are no gangs. But there's a mystery afoot, and Henry has to make choices — a la TellTale's offerings — while keeping in contact with fellow lookout Delilah over his walkie talkie.

Stunning for an unfinished product

Even though this preview is an early build of the game, things appear polished and every choice Henry makes is engaging. There's an artfulness here that's also pretty stunning for such an unfinished product. Firewatch doesn't have a release date yet, but we're looking forward to it dropping sometime this year.

While bits of the massively multiplayer game featuring space combat, mining, piracy, trading, first-person shooting and a persistent universe have been released, a bulk of the game remains to gamers little more than promise.

Meanwhile, Roberts Space Industries continues to prime interest in the game and feed the funding with a steady release of new ships sold like cars on a car lot for the in-progress game. Roberts says the company pulls in $3 million to $4 million a month and that 200 to 300 people a day buy the game.

So shortly into our meeting early this month, I ask Roberts what seems to me to be the obvious question: "Do you worry that Star Citizen could become a Ponzi scheme, that you're taking money for things now that you promise to produce down the line but never do?

"If the money stopped today would you be able to get all of the things out that you promised?"

Absolutely, says Roberts, creator of Wing Commander and Freelancer.

"For a start, people can only back for a ship when we have it in production," he says. "Right now there is a list. You can go onto the site and there is a whole bunch of ships that players know the name of and the stats of, but we haven't started the conceptual design phase, let alone the 3D modeling phase.

"It's not like we're selling stuff we still have to pay for down the road. By the time we end up like saying this week we're going to have the mining ship sale, it's like, we've already done a lot of the work on it and the remaining work is basically covered by what we bring in."

And those ships remain fertile ground for future funding.

"A bunch of people, they're not interested at all in combat, so they don't want that," he says. "They might be holding out for the mining ship. That will probably be, not hundreds of thousands of sales, but we tend to sell 7,000 to 8,000 of those ships.

"The ships are basically one of the core things that subsidize the building of the game," he says. "Definitely ships cost us money to make and a lot of our cost is a massive amount of people working on ships. But ships are one of the things helping us make the game."

So Roberts is very clear that the ships being sold on the site aren't made available until the company is sure they can be completed. But what about the game itself, the space those ships will fly though and the experiences they'll take part in?

To understand how Roberts is going to deliver such a vast game it helps to understand how the game is being rolled out. Star Citizen is being delivered in updates each building on the last, adding layer like a seemingly endless collection of Russian nesting dolls.

That first tiny doll given to players appeared to be a simple virtual hangar, but it was also the kernel of the game. Next came a ship inside the hangar and the ability to go inside the ship. Then came the first serious interactive part of the game: Arena Commander, which gave players the ability to hop in ships and get into space dogfights.

To get that playable element into the hands of gamers as soon as possible, RSI released it as a simulation within a simulation. That means the combat doesn't fit into any sort of persistent universe; damage taken, wins earned, losses accrued have no impact on the universe itself.

To make that work within the fiction of the game, RSI delivered Arena Commander as an experience accessed inside a holodeck inside the hangar, inside the game. As the game continues to grow, adding doll after doll, RSI will build on experiences separately until finally an update will tie everything together and set it live inside a persistent universe.

Roberts walked me through how that will all come about.

The stars and the moon

"Right now you can play Arena Commander, which is sort of the space combat and that is limited at the moment to the single sort of seeder ships and it plays very much like Wing Commander, like any space combat sim," he says. "This [month] we're going to release the first-person shooter part of the game to backers."

The shooter will initially launch with two modes. Roberts describes one of the modes as being sort of like Counter-Strike on a space station. The other, he says, is an Ender's Game sort of battle arena.

"So basically you can do all of those zero-g push-pull stuff," he says.

In April, a social aspect of the game will be introduced, he says. Once released, players will for the first time be able step outside their hangars.

"Then you will be able to open the door and step outside walk around and see those places," he says. "And then in the June, July timeframe we will release the multicrew part of Arena Commander, which means you'll be able to use those ships that can have more than one player in them. Basically, the biggest ship that you and three of your friends, you and ten of your friends fly now that will be in the backers hands. There will be a bunch of ships the backers have pledged for that are sitting in the hangars that they cant fly yet that they will be able to.

"Squadron 42 will be toward the end of the year. That's sort of basically Wing Commander single-player narrative story. And then at the very end of the year we will release the very early alpha of the persistent universe. It wont be nearly all of the systems and planets, but we plan to have five or six systems you can fly between. You won't be able to do all of the things we're planning on you to do, but probably trading, mining, piracy, combat and a lot of core stuff."

Then the company plans to spend 2016 filling out the rest of the star system, finishing ships, finishing characters "basically going from five to 130 star systems and adding more of the functionally and features on that we have and building out different roles."

"By the end of this year backers will have everything they originally pledged for plus a lot more," Roberts says. "But of course our intention is that it's a much bigger, more expansive, huger game than I ever considered we could do."

Virtually delivering the stars and the moons to players is no easy undertaking, and even with his grandiose plans, it far exceeds anything he originally thought he would be able to do with Star Citizen. So RSI has had to quickly build up staffing.

Thanks to Minecraft

There's been a lot of talk about the impact things like Kickstarter and Steam Early Access have had on game development, but neither of those things really had much to do with Star Citizen, Roberts tells me.

What made Star Citizen possible, or at least what pushed Roberts to give developing his dream game a try was Minecraft.

"The inspiration for what I did in Star Citizen was Minecraft," Roberts says. "Not necessarily because I was like, ‘Oh my god, look at the graphics.' It was the model that was used for development, this sort of intriguing basic game that would have never ever lived anywhere, no publisher would have ever agreed to back, but that came out of this grassroots community.

"Notch put it out there and basically says, ‘Can you give me' whatever he asked for and of course it was just him so he didn't need that much money, but he used the money to continue to add features and he was listening to the feedback to his game. He kept adding features, then he could afford to hire people and he sort of organically grew it out."

While RSI did eventually seek some funding through Kickstarter, that wasn't the initial plan.

"My original plan was that I was going to raise some money from private investors to build a sort of alpha that didn't have everything I wanted in it," he says. "It would have been enough that I could give it to someone and they could play it and they could give me a reduced amount of money and I would use that money to continue adding features until I built it to my final feature set."

In other words, the Minecraft system.

"That was my plan," Roberts says. "I was six months into planning that and had already lined up the investors and then Double Fine Adventure came along on Kickstarter and I looked at that and I was like, ‘Maybe I could advance that.'

"That was my sort of epiphany of, well maybe I don't have to wait until alpha, maybe I could get the people in sooner."

Even under this new plan of getting more seed money from Kickstarter and crowd-funding on the game's website, Roberts didn't think he'd get much money. Instead, he thought, he'd raise just enough cash to prove the level of interest to investors.

"I talked to the investor and says I wanted to do crowdfunding," he says. "It was a bit risky because if no one showed up they wouldn't have given me any money."

At most he hoped to bring in $2 million to $4 million with crowdfunding and add in another $10 million from investors to pay for a functional alpha. Then he planned to use that to start bringing in revenue which would be used to finance the rest of the game.

"It was 100 percent originally inspired by the organic growth Minecraft had," he says.

Instead, Star Citizens initial fundraising campaign brought in $6.2 million and nearly $40 million last year, astounding everyone, including Roberts.

Now, well into development on Star Citizen and with plenty of feedback from players, Roberts says he's happy going the route he did. A big issue with things like Steam Early Access, he says, is that despite how clear it might be that people are paying for an unfinished game, some players are still confused.

"Even on Steam Early Access you still get people thinking they're buying a final game and they kind of complain about it," he says.

Because the Star Citizen team spends so much time communicating with the community and because almost all updates talk about how early the game is in development, Roberts says they don't run into an overwhelming amount of complaints. And when those complaints come, it's usually the community that explains things to the players.

"I still see people on forums complaining about things broken ... and there's always a bunch of other people going, ‘It's an alpha, that's the definition of an alpha. You're hear to help make it better, not to necessarily play a finished game.'

That clear communication, even though it may boil down to semantics, is a big part of why the Star Citizen community is the way it is, Roberts says.

"It's the way you enter into any contract with anyone," he says. "It's about setting expectations. As long as you deliver on those expectations than it goes well.

"For us that's the nice thing about the relationship we have right now," he says. "The community feel like they are part of the development team and they give you feedback and they make the game better. Then we go back and reiterate and drop versions to them and they say, ‘This is balanced better' or ‘You fucked up the missiles now.'"

The development cycle of release, listen to fans, change and release again is bolstered by the game's tight release schedule.

"Every two weeks we do a patch," Roberts says. "Every two weeks we are dropping new functionality in or new balance in or fixing something. Sometimes a big update drops, like the first-person shooter update which includes a whole bunch of new content and features. And then there will be another patch two weeks after that and that might fix the bugs we didn't know about when we did the fist one."

The reason a game like Star Citizen might be able to release new content with bugs without spurring fan outrage is because of that relationship both developers and players agree they have, Roberts says.

"Halo: The Master Chief Collection's multiplayer is a complete disaster, right?" Roberts says. "We've definitely had situations where our multiplayer has been a complete disaster when we give it out to backers, but they're understanding because it's pre-alpha and they're like, trying to help you out so they're supportive.

"People are like, ‘Fuck you, mother-fucker.'"

"Once you move across that line into, here's the finished game pay me $60 for it and things don't work, people are like, ‘Fuck you, mother-fucker.' Which is kind of what you've seen this year, you've seen a bunch of games like that."

The reason Star Citizen doesn't run into that anger, Roberts says, is because of the understanding the developers have with the community. It's a different sort of relationship than what people purchasing Halo might have with those developers.

"You're paying the same amount of money, so it's 100 percent semantics," he says. "The thing that works really well for Star Citizen is that this is the contract that in their mind they made with us and it is the contract I feel they made with us. So they don't have an issue with that."

Those expectations are also a big part of how Roberts will decide when the perhaps perpetually-in-development game makes the leap from an alpha or beta to a game that can be sold as "complete."

"That's kind of my thinking of how I want to take Star Citizen across the finish line," he says. "When you go from people being backers to people being pure straight consumers, something changes in the way they view that relationship.

"I don't want to do that switch until I've had enough time with the backers to make the game as good as possible and everyone goes, ‘Yeah,  yeah, that's pretty good.'"

His hope is that the shift from beta to no-longer-in-beta will be, as with Minecraft, not that noticeable.

"That's what we think will happen towards the end of 2016 because that's when we think we will have finished the content and get most of the polishing in," he says. "We are saying before then we think it still be rough around the edges and have issues."

The most noticeable difference once the game is retail, he says, will likely be a change in the way the game will be priced.

"We will remove the crowdfunding aspects of it, and that will probably be the point where we do some proper marketing," he says. "At some point, once it is finished, we would probably do a push to get the general gamer in. Right now we're still hitting only a certain segment of the game population"

Galaxy Builders

Currently there are about 320 people, including contractors, working out of six studios developing the game, Roberts says.

"The way I look at the way we run the business now is we size our staff and what we're working on based on what we bring in every month," he says. "So if we're bringing a good amount, which we have been doing for quite awhile, then OK, I can afford to have a 300 person team working on it. If we didn't bring in the same amount it would have to be 200 or 150, which is still a lot."

Of the six studios working on the game currently, four are internal. Los Angeles is the corporate headquarters and is working on space combat, he says. A studio in Austin, Texas is primarily leading work on the persistent universe. A studio in Manchester, run by Roberts' brother, is responsible for Squadron 42 and helping with space combat. And a new studio in Frankfurt is working on the game's core technology, like its use of the Crytek engine, and is helping with the other elements of the game.

On top of the internal studios, Illfonic, based in Denver, is taking lead on the first-person shooter mechanics and Behaviour Interactive in Montreal is helping with operations, the persistent universe and is tasked with building a lot of environments with all of the different planets. That team is also woking on future concepts for the game which include iOS apps and things like Hololens, Roberts says.

On top of that, they have contractors working out of China and Mexico. Every studio helps out with building ships, he says.

"We are trying to split up each studio so they have a focus and lead they do so we can kind of work in parallel, so there is some crossover," he says. "We have about 200 or just under on staff and about 120 or so on contract.

"It is very big."

The globe-spanning mix of internal and external studios make management a challenge, but more concerning to Roberts is making sure that RSI's overhead never exceeds their income.

"That's why we keep a balance between contract and in-house," he says. "If we had to scale down you would be cutting down on the people on work for hire. I don't want to be in a situation where you have to scale down and people have to lose their jobs."

Finding all of those people, and making sure they had the right talents to help make the game what it needs to be, was helped along by some of the industry's bumpy year.

The relatively newly formed group in Frankfurt, for instance are made up of a lot people from Crytek, he says.

"Frankfurt is 17 now and it is probably going to be 40 in six to seven months," Roberts says. "A lot of them are core Crytek guys. Last year, Crytek has some financial issues and I think it spooked some of the people."

That combined with the developers wanting to work on the sort of game that would push the technology of the CryEngine helped RSI land a new team in Frankfurt.

"They're doing a lot of free to play stuff at Crytek, some of the engineers who built that engine were looking for a job," Robert says. "Epic was making them offers, id was making them offers and we were like, ‘Shit, this is the engine we're going to be using for years if we're successful, we can't lose that brain trust.' So we stepped in and said, ‘Don't go to Epic or go to id, we would be happy to set you up in Frankfurt and you can sort of work on Star Citizen.'"

The U.K. team in Manchester was formed with the help of Roberts' brother who was working at TT Fusions on Lego games.

Roberts says he and his brother had talked about working on Star Citizen from the beginning, but his brother had a solid job and Star Citizen was still a wild dream.

"I didn't want to say, ‘Hey, join me on this Kickstarter thing that I don't know if it is going to raise one, two whatever million dollars.' Once we got into the fundraising, though, it was pretty clear what the trajectory was.

"[My brother] says, ‘There are only so many Lego games you can do and this is the sort of game I want to be doing.' So we basically hired him and 40 people from that team, the core. They resigned and opened Foundry 42.

And with Foundry 42, Roberts didn't just land a team of developers who have worked together for years and led by his brother, a lot of those developers had a history making games similar to Star Citizen.

"A lot of the people working at 42 built Starlancer and Freelancer 2 a long time ago," Roberts says. "It's fun for them to come back to the genre and do that."

Making space combat cool again

Space combat seems to be cool again. The massive success of Star Citizen should be proof enough of that, but it's not the only proof. Eve Online, of course, has had a long running, if not mainstream, success as a space game. No Man's Sky has captured a lot of interest. And then there is the success of Elite: Dangerous.

When I asked Roberts if he thinks Star Citizen helped revitalize the genre of space combat, he declined to have the game take the credit, but did say that often all a genre needs to return is a bit of proof.

"I think it's one of those things that someone just had to do it and then everyone says people want this and then all of a sudden for whatever reason the prejudice changes," he says. "There was this prejudice that no one wanted these kind of games anymore, no one was interested. Finally, I think the thing Star Citizen did was show what you could do with today's technology.

"All of a sudden that changed the perception then I think other people feel safer committing to it," he says.

Roberts likens the shift to how people once felt about first-person shooters.

"It was all very sort of like Doom and Quake and fantastical and stuff and cool and fun and then there was Call of Duty," he says. "It was World War II and everyone was like, 'Aw yeah man, awesome, I remember shooting nazis in Wolfenstein. Fuck yeah, I want to go around in World War II. The first Call of Duty sold well but not 20 million units and it sort of built from there.

"But until that moment people weren't thinking of historical first-person shooters."

Another example, Roberts says, is World of Tanks.

"I swear to god you could have gone to any publisher and says 'I want to do a World War II tank game' and they would go, 'Well, do you want to sell 5,000 copies? We're not in that business get the fuck out of here.' Now World of Tanks has 40, 50 million people registered and they bring in huge amounts of money. And now you have Armored Warfare and everyone else going into the tank sim genre."

Too much stretching

With all of Star Citizen's new funding now coming to the company directly through its website, Roberts is in a strange position for a crowdfunded game: He's concerned about asking for more money for more features.

"We actually kind of backed off from doing stretch goals recently just because we've done so many of them and there is always a bit of a debate," he says. "First of all, we're starting to run out of ideas and second of all we have a lot to do and people started complaining."

So instead of stretch goals tied to new features, RSI is now offering up more behind-the-scenes looks at the game and its many parts.

"We shifted," he says. "Instead we'll do a deep dive every million; instead of a new feature, we're going to talk about how we're going to be doing something, the design, and reward people that way."

The studio has, in fact, already received funding through stretch goals of everything the team dreamed up before launching the project.

"A stretch goal was like if I drew out a big road map for this project, it's get out the base functionally and then go beyond what I wanted," he says. "Go beyond what I wanted is kind of where we are at now.

"We definitely have the funds, the resources to do that."

Moving forward, Roberts says, any sort of stretch goal would have to substantively add to the game.

And no wonder, look at just one place Star Citizen's lofty goals have it headed, future missions that are the ultimate expression of Roberts' grand vision for the action portion of the game.

"So for instance, one mission could be like, capture the Idris frigate," he says. "The first stage would be coming in and getting rid of fighter cover. Then you would need to shoot the defense guns off. And then you would go into the ship, board it and take it over. Basically, seamlessly going from flying around in space to getting aboard the ship and getting into a first-person shooter fight.

"That's the dream. That's the dream."