RPS Discusses: The Cloud

We recently sent Dan off to the Cloud Gaming Europe conference in London, where he interviewed Dave Perry. Following on from that he had a chat with Jim about this cloud gaming thing. This what was said.

Jim: Ok, so, you were at this Cloud Summit thing, eh? Who was there? Why were YOU there?

Dan: Ha, I was there because a) I got a free press pass and someone told me there were going to be sandwiches b) because we're often told that streaming / cloud technology is taking over and I wanted to hear the arguments of the interested parties and c) David Perry was there and you told me to go.

Jim: Oh yes, I did. Well, I hope it was interesting. Who else was there? What were people saying?

Dan: Well, it was interesting. "cloud" gaming is a bit of an ill-defined subject - so the CEO of Eve was there, saying he considers Eve a cloud gaming platform and everyone was like "okay, I guess." The real stars of cloud gaming though are the streaming technology guys - their tech sort of democratises the cloud, making it accessible to all of us gamers. What they weren't saying was that they were going to take over all desktop gaming - but there's certainly an argument that they're capable of doing it.

Jim: It sort of gives out readers the fear, doesn't it, the idea that computation is being taken out of their hands/desktops? Why do you think that is?

Dan: I mean, you and I, consider always-on, infinitely-powerful tech as part of our science fiction birthright. This stuff is coming, right? It just seems that there are barriers in the way. Why are they scared? I think they have fears about the loss of agency and what a large corporation could do if it got control over the way we play games. A corporation other than Steam, that is. Do you agree with that?

Jim: Well any corporation controlling the majority share of a platform is dangerous, I think. The value of the gaming web right now is its democracy: anyone can do (almost) anything, because everyone owns a computational device of lesser or greater capacity.

Dan: It should be something we're excited about. Always-on access to all our games, where ever we are, at quality higher than we've ever seen before?

Jim: The idea of gaming being locked up in the server-farms of just a couple of companies gives me the willies, frankly.

Dan: Yes; but what about a distributed cloud, instead? Where gamers sell their spare capacity back to a clearing house, that uses it to supply people with remote processing power - a bit like the national grid and solar panels? Is that pie in the sky do you think?

Jim: Well that's the interesting issue, isn't it? Can the web maintain its sort of "everyone is a node" philosophy now that it is about big money? Commerce vs what would be ethically and culturally valuable.

Dan: It's hard to say; the question is whether any big corporation would invest money in something they don't own wholly; also, given that the point of the cloud for the big companies (Oracle, Cisco), etc, is so that they can make more efficient use of their computing power, leverage the difference in cost between a home computer and a rented slot on a server farm to make money... my pie-in-the-sky idea would only work if state or inter-state led. And they've got bigger things to worry about at the moment.

Jim: I don't know if it's pie in the sky, though. I think if the services appeared for genuine distributed processing systems to work, people would use them. I suppose what interests me about the cloud is that it is revealing that actually computers were about communication, rather than computation. There's so much computation now that it's sort of incidental, and the real issue is: can you transmit the data?

Dan: Yes; again, I got to interview Ray Kurzweil yesterday, and he's predicting the advance of computational power to increase so ridiculously quickly over the next ten years, that'll be hundreds of times faster than right now. At the moment, there's a bottleneck in the data. Our national networks aren't fast enough to handle the data reliably. They don't stay up, bandwidth is an issue outside of conurbations, and latency is a huge issue. It makes you feel like the data-supply side of the cloud needs to be fixed before we can start taking this seriously.

Jim: Yeah, and all that really matters now is moving the information around. I was reading a Brian Aldiss book at the weekend.

Dan: Oh, which one?

Jim: I've mentioned it before, I think, it's called "The State Of Further Things" or something like that, and it's a bunch of essays written in a cottage outside Oxford in 1970.

Dan: Ha; I wonder if Theodore Sturgeon was still alive around the corner? Carry on. (I mean, Olaf Stapledon. Duh, Dan.)

Jim: Aldiss talks about wrist-mounted talkotrons (phones) and "The Big Hookup" (the internet) and it's a remarkable sort of vision of the net that gets a lot right, but cloud gaming is the bit he couldn't have predicted: that the access to remote information could be so fast and so much that you could transmit real-time rendering of something that was only vestigially imagined in the 70s.

Dan: Yes; it is striking in those old science-fiction films that computers are almost never remote; they're present and the most amazing thing they can do is produce a render of a human face (e.g. Red Dwarf...)

Jim: I wonder if that's part of the issue with Cloud gaming, it's not just the slightly muddy gfx, the latency, the ownership, it's facing something unlike how we understand computing to be.

Dan: I think that was the interesting thing about talking to Dave Perry; was the slow realisation that this technology can be used for almost anything, but it's very good for two things: Data security. And data projection. Hence, the US military wanting to use it to stop Wikileaks. So, that analogy I made before, to us coming full circle; we went from mainframes with remote terminals and we might be going back to mainframes with remote terminals. The distance has just increased. I think you called it "computing as industrial process".

Jim: Yes, which it was originally, until the 1960s. There's this great bit of narrative in What The Dormouse Said, which is about the PC revolution in Silicon Valley, where this guy calls computers "personal idea amplifiers" - again, amplifier/transmission, etc - and it was this huge leap to computers not being vast industrial number crunching machines, but personal and personalised devices, and now it seems almost like we've gone through the circle of coming back to that, only what is industrialised is the personal idea amplification.

Dan: Ha, yes. It's communication again. Yep. It's an odd one, because the small devices are so powerful and cheap to make, that to turn them back into terminals seems pointless - but the additional computing power drawn from the cloud is so huge... it's just a matter of what we can possibly do with it. The whole concept of terminals in Iain Banks' books - the earring that allows access to the computing power of an orbital. The terrifying thing is we're so close to that already. This could just be the next step. Powerful devices with access to HUGE amounts of remote power. The question then is; what do we do with it? I mean... programming games or applications for an infinitely big computer? Obviously, the limits are still the cost of employing all those computers, but you can imagine people making arbitrary separations.

Jim: One of the issues here is that this stuff gets beyond the capacity of people to see the full extent of uses for, especially individuals thinking about personal needs.

Dan: Yes; which again, is odd, given how software development, in only the last few years, has become totally democratised.

Jim: But yeah, there's a cost thing, too. I mean it's very hard to be sure that cloud computing is going to be cost-effective within gaming. For other stuff, definitely, but when you are suddenly running high end GFX cards by the thousand... I think the thing we were discussing was how, or whether, it could be a system that evolved like the consoles or PC have. I mean what is the "console cycle" for a set of remote server banks?

Dan: Yes; that's a really interesting question. And what is a console if there's no limiting hardware save for the input peripheral? Will the next Wii just be the Wiimote? The next 360 just the kinect?

Jim: It has to deal with these kinds of issues, and I expect it will. I think Gaikai/OnLive is like the 3DO console: making the leap into 3D, texture mapped rendering on your TV, but failing. Same thing with streaming and the cloud.

Dan: The specifications games will be developed for won't be cpu, ram, etc - they'll be amount CPU-load per minute, etc.

Jim: In a couple of years there will be a revolution akin to Sony launching the first Playstation, and Cloud gaming will find its place. (I think a lot of gamers forget the transition that the PS1 represented, actually, it was such a big deal. Way beyond the 16-bit era we all so fondly remember.)

Dan: That does make sense, yes. This is like the early days of tablet PCs and Mira / smart displays.

Jim: But I don't think it's here yet. What was the sort of consensus at that conference?

Dan: I think, possibly, only for demos. Which have interesting implications for themselves. E.g. if anyone can try any game for a limited time period, it erodes the influence of marketing. Sure, I'll try EA's Army of Doo first, but then I'll try all the indie games too. Imagine a F2P streamed game that charged you 10p an hour, perhaps only after you've played 20 hours. it also opens up bizarre new business models.

Jim: Yes, it's a big shift, and it's the one that ends having to go to the shop or wait for a download. Which actually, when you see someone use OnLive for the first time and they go "oh" because it just plays instantly, there's this shift in comprehension. Well, not so new - do you remember Wild Tangent?

Dan: I showed OnLive on my iPad to a random celeb when I was at Toy Fair earlier in the week and his eyes just boggled. WildTangent? No, what was that?

Jim: A web game company that went bust, but threw out some great ideas. Years ago they were talking about paying per-minute on games. They were all "well, if most people only play 2 hours of Crysis why should they pay the full $40, would MORE money be made if the games industry actually catered to the 2hour players". Now that's actually possible in a way it wasn't just a few years ago.

Dan: Sounds like they were ahead of their time - but they won't be charging for the first 2 hours, but when you're already totally committed.

Jim: Yes, but you are right, but it opens up a very different set of possibilities for what you pay for.

Dan: So, at the conference, I think pretty much everyone was confused. There are all these cloud suppliers, there's OnLive and Gaikai, and no-one knows what to do with any of them. Making the jump just seems terrifying, especially to companies that have only ever made console games.

Jim: Indeed, and those larger companies are hilariously blinkered, as the Larian interview this week demonstrated. Which could mean we get a wave of genuine innovator type companies swooping in and fucking everything up.

Dan: Yes, the Zynga for the cloud is out there somewhere, gestating...

Jim: Eugh! That's what makes people worried.

Dan: It's made of clouds! It's a big cloud baby, coming to eat up all our PCs.

Jim: Zynga is eating itself, anyway.

Dan: Yeah, but they held off the collapse until they went public. Clever girls.

Jim: Yeah, it was interesting to watch. But there was no hiding the horrible, cynical models it worked with. That's games as industrial process.

Dan: Oh, god, yes. The behavioural-shaping stuff. The thing was, those players who got hooked and then put off - they won't come back to games for years. I mean, does anyone play Facebook games anymore? They've disappeared off my pages.

Jim: Yes, I am sure there are still millions of them, but anyone who sinks a thousand hours into that will come away empty, and they know it.

Dan: Yep, empty, poor and feeling exploited.

Jim: It's a horrible impression of what games are or could be.

Dan: Yes. Like Shockheaded Peter, trapped beneath the floorboards. *Shudder*

Jim: Ok, let's wrap this up.

Dan: Okay! Have we learned anything?

Jim: What's your feeling about the cloud and its potential impact on gaming?

Dan: I think that we'll definitely see it being used to demo games more directly - the banner ads at the top of this page will almost certainly be one-click Gaikai demos at some point, though I'm not sure that'll last. I think publishers will have those streamed demos on their home pages, but I'm not sure retailers will buy into the tech. I think PC gamers will only move across if the quality of games matches the state-of-the-art on PC, it's reliable, it's cheap and the problems with bandwidth/latency are sorted out. I think one of the next consoles will have it built in - probably the 360. (But definitely not the WiiU.)

Jim: Yes, it seems years away from being "fully-fledged". And anyway it's just another channel. Albeit one that contains traces of the future, or at least more traces of the future that a lot of what people are offering currently!

Dan: Yep; the key thing we learnt from books > radio > TV > computers is that channels don't disappear. They just layer up and we use them all. Advances in tech will still be needed, whether to power home computers or server farms, so the stagnation we're seeing at the moment, from development resources being diverted to tablet development, is only temporary.

Jim: Hurray for technology!

Dan: God, I love living in the future. I keep expecting Judge Dredd to ride by at any moment.

Jim: Oh he'll be along. He's just getting David Cameron to sign away your human rights.

Dan: Hurray for the future! Hurray for Judge Cameron!