Ascent: The Space Game is one of the MMO genre’s most ambitious titles. Whether it’s the seamless transitions between the vastness of space and the vastness of its planetary environments, or whether it’s the emergent gameplay made possible by Ascent’s NPC AI, this particular PvE sandbox is doing things that other games can’t (or won’t).
MassivelyOP recently interviewed Fluffy Kitten Studios founder James Hicks about that emergent AI and several related topics. Click past the cut to learn how this tiny indie studio is shaming its big-budget counterparts and assembling a feature-rich title on a shoestring budget.
Ascent recently made news due to its President of the Galaxy election, and a press release noted “a direct line to the developers” for the election winner as well as the ability to “influence the foundation of the universe.” Do you foresee player involvement growing beyond a single individual in the future, possibly to something approximating EVE’s CSM or SWG’s player senators?
There are a few ways I interact with players on a daily basis. Forums, Steam forums, in game chat, and we even have an IRC channel. Giving the President the “direct line” means there is one conduit where I know the feedback is representative in some form. Usually it’s representative of the “old, established players” crowd, but this is the most informed section of our playerbase. Both the outgoing and incoming president have a history of discussing things with senators and obtaining consensus, so they can function as a spokesperson for large blocks of the playerbase.
Of course, their main role is in-game. The Colonial President is the only player who can approve player made gates connecting to the inner systems, effectively expanding the newbie zone.
Presidents also have an ongoing roleplayed dialogue with the United Nations Combined Arms, the main NPC faction in the game providing security for colonies and inner systems. We discuss capital ship deployments for security, and the President has just this morning appointed an in-game reporter who can contact UNCA with questions. I think that’s a better way for the game’s story to unfold – just publishing stories means 1% of the playerbase might read it, but a question and answer interview format with contentious questions asked by an appointed presidential reporter might stimulate a lot more interest. At present the playerbase are quite suspicious of UNCA – they’re a large and very powerful organisation whose motives are largely unknown.
Ascent would be an ambitious title in terms of feature sets and functionality for a large studio in 2015, never mind a small one like Fluffy Kitten. Why do you think it is that MMOs in general have gotten simpler and gameplay options less diverse, even though budgets and team sizes have gotten much larger? Do you have any plans to grow the studio beyond your current staff, or are you happy with the current development pace and the associated reaction of Ascent’s player base?
When you’re a manager in a large corporation you spend a lot of your day eliminating risk. I know, I used to do it. If one of my staff had come to me with the Ascent feature list and said “We want to develop this!!!” I would have immediately said “That’s a crazy mountain of technical risk! I’ll never get funding for something like that!” and it would have been absolutely true.
The more money involved in a project, the less tolerance for risk there is. The stakes are just too high. As AAA MMOs get bigger and bigger budgets, frankly, we can expect them to get less and less innovative, in general. I hope one day that isn’t true but right now in 2015 it’s the case. So, weirdly, if you want to try some insane things, like 200,000 structures in a player-made city, getting out of your ship and walking around on earth-sized planets, or open-ended player colony, star base and jump gate building, your only chance is a small company or basically to do it yourself. I remember when I told my old boss what I was doing and described Ascent, he was interested in investing. I had to tell him straight up: “Right now I think this project’s far too risky to invest serious money in.” And I meant it!
The Ascent project for me is all about going right ahead and DOING all of those things I know anybody who’s accountable to shareholders has the good sense not to touch. There are solid reasons people don’t try these things. Huge numbers of structures leads to huge CPU, RAM, and GPU usage, even with multi-threading, every trick in the book and some new ones.
Every little mistake I make results in memory leaks that cause crashes in minutes, sometimes seconds. An Earth-sized planet, in any modern game engine is a complete nightmare to work with. Modern game engines can’t really cope with the distances involved, especially not when you place something human-sized on the surface and want it to walk around. I’ve spent many many hours working around these problems, with no guarantee of success, at each stage of development. We get them sorted, but every time we add something new to the game they come up again and need a whole new level of workarounds.
For example, at first, walking a human-sized avatar into a three story building on the surface of one of these planets, at a colony with over a hundred thousand other structures, just didn’t work. Everything from the camera not scaling down it’s movement properly from viewing space ships hundreds of meters long, to fitting into a 10×10 meter room, to the avatar falling into floors or walls because the game engine couldn’t cope with the scale or position a building 6300km from its parent object properly etc. Game engines use 32 bit floating point for positioning and it’s just not accurate enough for that kind of scale. All of these issues had to be fixed – weeks of work – before we could reliably do something I can achieve in literally minutes with a simpler game: just walk around a room with a human avatar.
So while the technical risks are huge and I pay for it with many extra hours of work to do the simplest of things, the end result is: we have a game where you can experience some totally unique stuff. And that makes it all worthwhile. The first time I looked out of my avatar’s eyes, out the window of a building onto the colony I began building over a year ago, seeing it from that perspective, with the barren, alien terrain in the background, the ocean, the clouds, the sun setting, it was a really emotional moment. I can’t describe it, but it was one of those times where you go “We really are onto something unique here!”
As for growing the studio, I would love to, I just have to keep an eye on the budget. At the moment based on sales it looks like George (our artist) and I can make it to a full release sometime around December. George has made leaps and bounds with the game’s visuals – I recently had to replace all of the screen shots and now most of our videos are hopelessly out of date too (check out the new jump effects!) – but neither of us are really UI experts, and I think the UI, despite looking a LOT better than it used to now, really needs work in the usability department. It would be great to hire in that space but I’m holding off for now to see what sales do.
The vast majority of our potential player base has never heard of us. MassivelyOP is the only major site that’s covered us so far. We’ve had strong enough sales to get by ANYWAY, which I take as a good sign if we ever do break out.
I would like to develop the game a bit faster and a lot further. I’ve got experience running a department of up to 18 staff and I think we could easily add a few more people and speed up a bit. But that all depends on sales as we have no external funding.
Have you kept track of other attempts at emergent NPC AI in the MMO space (something like the now-defunct Storybricks, for example)? There doesn’t seem to be a lot of interest among developers to move past the MMO AI status quo aside from Ascent and a tiny handful of other projects. Do you think that’s because of the costs involved or is it more of a programming/skills challenge?
In general I think AI is a really expensive way to solve a “content” problem. If you can throw a story script and some art at it, that’s going to be the shortest path if you’re a big studio with writers and artists on staff. Worse, for an MMO, AI inevitably means CPU gets used up server side. There’s no way to get completely around that. So it’s generally to be avoided. In fact, when you say “NPC AI in the MMO space” I pretty much picture dusty plain and some tumbleweed. All of the really exciting NPC AI I know of is happening in small Indy games like Dwarf Fortress and the next MMO Tycoon. I wasn’t aware of Storybricks until you mentioned it – but I looked it up, and I think they were really onto something there. However, predictably, they don’t seem to have been able to sell investors on the idea. I couldn’t either.
But if you’re tiny like us, AI starts to make sense. I’ve got one artist, no writers, and nobody who writes quests (other than myself), so we rely heavily on procedural quest generation already. When I write a new storyline quest, I define a new quest TYPE and that sort of quest shows up everywhere. AI gives me a similar ability. It’s a lot of work to get something very basic going, but with thousands of individual NPCs waking up, the opportunity for procedural gameplay through individual AI is quite real.
It’s a very high technical risk approach that I could never sell to investors, but the result is something unique – we already have NPCs flying ships for players, transporting goods from A to B (and to C and to D and to E, etc.) for profit, all hired and managed through interactive text chat. Nobody’s done that before. And for the server-side CPU, I spent fifteen years in corporate IT figuring out how to deal with big data and big processing loads – for a tiny company we’re surprisingly well placed to take this on.
Can you walk us through a basic quest in Ascent as it will function once all of the NPC AI stuff is finished and implemented? A recent press release mentioned that NPCs could get addicted to substances and attempt to fill their own wants and needs, but how exactly will that manifest from the perspective of an Ascent player?
The next major feature we’re implementing is smuggling. There are two sides to it, one for colony owners, and that’s quite complex, and one for smugglers, who could even be a new player on their first day, and that ranges from very simple to extremely complex.
For a colony owner, we’ve recently seen the introduction of a new asteroid type that yields neurocrystals. These can be processed using new technology into a variety of products, some of them being designed as neurostimulants. We’ve also added alcohol products as a research item, and tobacco fertility has shown up on a number of planets. These industries all produce goods that not every colony owner will want to import or trade with.
So we’ll be adding a simple system of Colonial Laws, initially just dealing with what goods are disallowed for trading at each colony or outer star base. So you can, for example, ban cigars. Later we’ll expand this feature into a whole host of other laws that influence culture over time and make each colony unique in its own way. This is important because colonies and outer star bases are the adventuring playgrounds of several upcoming features. Smuggling leads pretty directly to assassination and bounty hunting – the informant, the customs official who wont take a bribe, the cigar dealer ruining your colony, all targets! Hunting down individuals in vast, unique, player made cities… will be nearly impossible to get working but once it works, it will be another totally unique experience.
Meanwhile, other colony owners will be producing and exporting these goods as a legal trade. The stock market at a colony where they are banned wont take them, so how do you smuggle them in? You will need to establish an NPC contact willing to take your goods and sell them for you. For a small amount of goods – up to a few tons perhaps – this shouldn’t be too difficult. Talk to enough people and you’ll hear who buys that sort of thing, or run into someone who’s interested in starting out. But to distribute thousands of tons you’ll need a vast network, and that will take a lot of time and investment to set up. As yet, I’m not sure how much of that effort will be made by the NPCs themselves. It depends on the level of AI I can get together. And then there’s muscling in on a network someone else set up… but let’s not get into that yet.
So if you’re a new player and you want to start off smuggling, you buy a few tons of cigars some place that makes them, and go talk to some NPCs at a colony where they’re banned. Ask enough people “hey, do you know who buys cigars around here?” and you’ll find some people who will take a ton or two. Once you’ve got those contacts, they’ll buy from you again tomorrow, and likely in growing amounts as they expand their own networks and customer base. But if you’re an established player, playing your version of “Smuggler Tycoon” and you’ve landed with 60,000 tons of Cigars, you’re going to need to stitch a bunch of these NPCs together into a vast criminal network.
If you’re selling cigars or beer at colonies they’re banned on, you’ll likely find a lot of people willing to buy, comparatively. Whereas, neurostimulants are a lot more dangerous and more of the population frowns on their use. You’ll have a harder time selling and putting together a network, but the profits could be much higher.
But just finding buyers isn’t the whole story. You also need to get the goods past customs. Which means I have to talk briefly about colony owners again.
I’ve got some simple rules where one player’s actions can impact another player’s investments in the game:
- The action has to be limited in its impact. If you flood my colony with every illicit thing under the sun, most of my colonists will have enough common sense and moderation to get by. Some won’t. It’ll impact morale, health, and productivity. But not enough to make me rage quit. Morale, health and productivity are in-game already, so impacting them with another factor isn’t a big step for us.
- The impacted player needs ways they can mitigate the impact. Customs ships and officers scanning cargo. Research into medical treatments to mitigate the health and addiction impacts.
- The mitigations should be on a sliding scale. Super high tech you poured everything into for months might be able to make the world’s worst Neurostimulant victim a productive, happy, healthy member of society. But the first quick level of research will take the edge off the problem. So if a player cares a LOT about this issue, they can choose to do a LOT about it, but someone who finds it rather tedious can reduce the impact to a level they can live with and move onto something fun. Best of all colonies are unique not just in what products they’ve banned but how they approach the problem of smuggling and its impacts.
So evading and/or bribing customs officers will be a thing. How much of a thing depends on the AI and the level of implementation I can squeeze into this release. We do have a good radar and terrain system in game already, as well as under water activities and cloaking devices, so it’s not like we’re starting from zero. I think there’s potential to make the physical act of smuggling goods quite a serious piece of gameplay on its own. The natural market effect of supply and demand will mean that the best protected colonies pay the most for their banned goods.
So you can see, actual AI where you can hold a conversation with an NPC in text chat and they can make a decision based on their personality and information available to them… lets us leave the concept of the traditional quest behind. Standard quests / contracts may or may not form part of this release, but they will make a comeback with bounty hunting and assassination as these fall into pretty straightforward “quest” bounds with one achievable goal.
Yes, with some provisos. You can do this to a limited extent already with NPC traders and colonies. Colonies produce a range of goods 24×7, as long as the colonists are happy, healthy, and fed. Which goods are produced depends on what resources are available on the planet you settle on, and what concentrations of minerals and crop fertilities (if any) are there. NPC traders can now transport goods from any colony or Outer Star Base to any other colony or OSB in the game – so long as they can fly directly there via gates or within the same star system. They’re not fast as they can’t warp or use hyperdrives, but they get the job done.
But there always has to be a limit. You can make serious money right now with NPC traders, trading between other players’ colonies – especially if you pay attention to the stock markets and what’s available. But they can’t haul goods into the inner systems and sell them to NPC stations for credits. Only players can bring new credits into the game manually by doing that (and other things which pay out credits). In addition, NPC traders are required by law to only trade up to 50% of available buy and sell orders in a 24 hour period. This means that at any given time, half the market at colonies and outer star bases is reserved for players. Over time this equalises with one full day’s production in storage reserved for players, and 100% daily production moved by NPCs – until a player comes along and trades.
Similarly, while colonies can produce a lot of stuff, they can’t produce everything. Many elements required for ship, structure and module construction are only produced by mining asteroids. That’s another thing that only players can do as it’s a manual, interactive process – beaming a moving target up to the right temperature.
I’m unlikely to ever let NPCs cross these barriers – this keeps the economy under a certain amount of control, by tying the money supply directly to player population and activity. It also keeps inflation down to a reasonable level, and ensures that new players ALWAYS have a market for their goods and services.
You mention companions though. When I play Wurm Online one of the things that keeps me playing is the strange connection I feel to my horses in the game. True story! All they do is eat grass and run around, but I raised them… and their parents, and their grandparents, and their great-grand parents and all of that investment means I feel attached to them as beings.
If horses with no communication, no personality and just an eating and running animation can do that, what can NPCs do that are programmed to talk, think, feel, have ambitions and personalities? At present they spend a few minutes as members of your crew while you move them to their new ship. As we move along, what will they be like as smuggling contacts, ship crew, employees, rivals, friends, family…?
Thanks very much for your time!