User:Juval/Test2

KAIZEN IAE ANTICIPATION
Done. Ref: Prasad Naturals lawsuit.

Open: Third quartal 2949: "Anvil Aerospace quarterly projections predict a downward turn after a a series of supply chain disruptions forced them to temporarily shutter several manufacturing facilities in Castra."

PORTFOLIO: YORM
-#seo: - -TNT|Infobox Commlink -
 * description=Star Citizen - Comm-Link Archive - "Few component manufacturers have a more colorful backstory than Yorm."
 * Title = Porfolio - Yorm
 * Image =
 * Source = https://robertsspaceindustries.com/comm-link/spectrum-dispatch/17339-Portfolio-Yorm
 * Type = Spectrum Dispatch
 * Publication Date= 2019-11-06
 * Series= Kaizen

Few component manufacturers have a more colorful backstory than Yorm. Named for the famed racing pilot Gotlieb Yorm, the company brought his revolutionary shield design to the masses in 2822 and successfully spun it into a manufacturing empire. Today, Yorm remains one of the best-known manufacturers of competition-grade components in the UEE. At least part of this fame is because it embodies the colorful, eccentric and outlandish personality that made the pilot famous.

The Iconoclast
No one knows where Gotlieb Yorm was born or raised. As surprising as this may be in our day and age, biographers, investigative journalists and fanatical fans have all tried to uncover his origins, but failed. This lack of hard evidence led to the popular theory that he was raised as an indentured servant in a Banu ship souli. For many, that theory would explain his unique skill set, deep understanding of ship design and complete disregard for anything that came before. Gotlieb refused to comment on his past, preferring to remind people that, “my ship doesn’t have a rearview mirror, so why should I?”

Gotlieb first made a name for himself in 2796 on the underground racing circuit in Baker. Piloting an old Aurora, he won his first race in commanding fashion and immediately garnered attention from racing enthusiasts for his eccentric and outgoing personality. Whenever asked what he hoped to achieve, his answer always remained the same — to win the Murray Cup.

Adel Fansekar only needed to watch Gotlieb race once to know that he was the real deal. Famous for her eye for talent as well as a massive bankroll, Adel had backed successful racing teams for decades, yet none had achieved the ultimate goal of winning the Murray Cup. Gotlieb turned down her initial offer to finance a racing team, claiming that he was doing just fine on his own. His blunt dismissal shocked Adel, who was convinced it was a negotiating tactic and returned with an even larger incentives package. When Gotlieb rejected the second offer, Adel asked him point blank what it would take to work with him. The answer was simple: autonomy. Gotlieb supposedly said, “I’ll always take your money, but never your advice,” and with those words their partnership was sealed.

Properly financed and in full control of his team, Gotlieb quickly earned a reputation as a racer who would do anything to shave a few seconds off his lap time. Race aficionados reveled at his unusual ship modifications and flair for dramatic finishes. Yet for years, a Murray Cup championship stayed elusively out of reach.

Gotlieb’s popularity, however, grew. He became a household name in 2814 when he began racing completely nude, claiming that clothes were nothing more than unnecessary weight. It wasn’t long before the Safety Commission intervened and required all racers to wear flight suits. In search of another advantage, Gotlieb focused on his ship’s shield, since his strongest Murray Cup race was Blitz, which allows combat up to the point where an opponent’s ship becomes disabled but not destroyed. After losing the 2816 Murray Cup Blitz title by less than a second, he famously ripped off anything he deemed unnecessary from the shield generator. When his chief designer noted that the shield could now only fend off a few shots, Gotlieb replied that’s all it needed to do. If he couldn’t navigate out of targeting range within a few shots, then he’d lose the race anyways.

Little did Gotlieb know that this decision would not only revolutionize racing, but pave the way to achieving his dream.

Less is More
Racing with his new streamlined shield, Gotlieb Yorm finally won the 2817 Murray Cup Blitz. Meanwhile, Adel was both ecstatic and observant of the pilot’s popularity. She recognized the desire of other racers and ship enthusiasts to embrace Gotlieb’s “Less is More” philosophy. She approached Gotlieb about mass marketing his shield. He agreed under three conditions. First, that Adel manufacture the component herself and not license it to another company that probably wouldn’t do it justice. Second, that the shield not reach the market for five years, so he’d maintain his competitive edge and have a chance to develop newer designs for his ship. Finally, that all he’d have to do was lift a pen to sign the agreement, then sit back to collect the royalties.

Adel agreed again to Gotlieb’s terms. She later claimed that this five-year grace period was a blessing in disguise. She’d financed and run multiple racing teams, but never delved into component manufacturing. For years, Adel kept her plans quiet while undertaking copious research on the industry. Slowly, she poached respected managers and consultants to execute her vision and made them sign an NDA that stayed in effect until the product launched.

Adel’s meticulous planning led to Yorm’s first shield coming off the production line a little over four years after her deal with Gotlieb. Though the pilot had insisted that he didn’t want to be involved past providing the initial shield design, she wanted to give him the chance to weigh in on a product that would bear his name.

To Adel’s surprise, Gotlieb became engaged in the process. He made a handful of valuable suggestions based on his experience, including advocating for the use of more expensive, but lighter, materials for the piping system. When asked what he thought about various aspects of the component, Gotlieb would always respond the same when he was ultimately satisfied: drawing a giant happy face. Adel liked his response so much that she asked him if they could use it. It’s been the company’s iconic logo ever since.

Don’t Worry
Yorm’s first shield generators became commercially available in 2822. Propelled by the cult of personality around Gotlieb and the fact that the shield was his exact, approved design, demand for the shield was astronomical. A lot of other companies might’ve been overwhelmed with such a popular debut product, but Adel’s careful planning meant that they were ready. Yorm shield generators had been coming off production lines for close to six months and could be hauled wherever needed.

Adel carefully scaled Yorm over the following decades to manufacture a wide range of competition-grade components. Though Gotlieb never personally designed or even consulted on any of their subsequent products, all were built following his “Less is More” philosophy. Adel’s hard work and vision turned Yorm into the successful company it is today.

Yet, the journey was not without its mistakes. For a brief period in the early 30th century, the company, under the leadership of Silvio Halbrook, attempted to expand its reach into hull plating, but a series of bad design decisions, mismanagement and manufacturing issues almost sank the company. After Silvio’s ouster, Bao Ingram took the reins and returned Yorm to its roots, competition-grade components.

Yorm has returned to its previous glory under Bao’s control. It remains popular today among racers and those who prioritize performance over all else. Those that love the company claim its logo perfectly represents how they feel when flying with Yorm components.

= CitizenCon 2949 - Building a Dynamic Universe =

[Introduction]
"Thus far we've been focused entirely on a single system: Stanton (...) four planets, twelve moons, 225 economic nodes, 44 trading outposts, 13 Rest Stops and 50 asteroid fields."

- Tony Z

TZ: (...) Hi, I'm Tony Zurovec, I'm the Director of Persistent Universe for Star Citizen. (...) and over here on the right on the keyboards is Jake Muehle, our economy designer from Austin. (...) We're going to show you some interesting things tonight and I want to make sure to call out our guitar and bass players back in Montreal, Martin and Rivera Lizey [?]. They spent a lot of time working with us over the last 4,5,6 months very intensely on what were gonna revealing tonight. (...) You need some background information to truly appreciate it (...) and you'll see why we consider it a quantum leap.

So as everybody here obviously knows star citizen is a very ambitious project, our goal is to deliver a variety of incredibly detailed solar systems that behave in logical fashion and that they can give you ample freedom to follow your interests whatever they may be. Thus far we've been focused entirely on a single system, Stanton, which includes four planets, twelve moons, 225 economic nodes, 44 trading outposts, 13 Rest Stops and 50 asteroid fields. (...) The point of this presentation then is to give you a better understanding of how, when the first one's taken so long we're aiming to kick things into overdrive so that we can deliver new systems much more quickly. This involves obviously a lot of different areas: procedural generation of terrain, modular art sets, algorithmic detailing of buildings and interiors and a lot of other stuff.

[Dynamic Content]
"Dynamic Content responds to player and NPC actions in logical fashion, its systemic. This means that the universe is in constant motion which makes it a far more interesting place to explore."

- Tony Z

For for the purposes of this talk though I'm gonna focus on what I call the Dynamic Content. So what is Dynamic Content ? Dynamic Content responds to player and NPC actions in logical fashion, its systemic. This means that the universe is in constant motion which makes it a far more interesting place to explore. Most missions should of course be dynamic in nature. They should only be offered when there is an actual need for something. This might take the form of a factory that needs some aluminum to create a retail product, a criminal that needs to be brought to justice or someone that keeping them from one location to another. At their core, missions or services provided by one party for the benefit of another and they are therefore one of the fundamental ways by which imbalances in the economy can be addressed. Likewise, economic nodes which include things like refineries, factories and retail shops need to be dynamic, since they must continually take into account the various market forces and adjust accordingly. When demand for something outstrip supply. Whether it's iron ore, bullets or ship repairs you expect prices to rise and conversely when there's an excess of something you expect that prices will fall. There's more to it of course. Concepts like marginal demand and price discovery require a fair bit of math, but in the end it's all about the constant search for equilibrium as production and consumption levels vary.

I'll get around to explaining probability volumes in more detail in a bit, but suffice to say for now that they dictate what you encounter as you travel through a given area. As such they should definitely be dynamic so, i.e. if a new field of asteroids, rich and valuable commodities is discovered you'd expect to find a lot of NPC miners in the area, but as the resources are gradually depleted, their numbers should start to fall off. Pirates should gravitate toward areas with a lot of value to plunder and recoil as the amount of security ticks up. So let's look into each of these areas in a bit more detail and see why we've struggled to set up a single solar system and at the same time get a precise understanding of what needs to happen, so that we can start moving a lot faster.

[Missions and Mission Design]
We'll start with the missions. A Star Citizen mission is fundamentally just a container for code and data similar in concept to a class in any of a multitude of modern a programming languages. Inputs can be passed into a mission from either the runtime environment or services to allow for customization. This allows designers to do things like say inject the type and quantity of cargo whether the ship should be damaged or not where the ship should be heading into a generic transport scenario so that we can reuse the same framework for a lot of different situations. Missions can have content tags embedded within themselves to denote their composition or whether they contain, security, pirates, freighters, miners, asteroids or some kind of accommodation. Missions can also have somatic tags applied. These don't specify physical content but rather detail the headlines story behind the scenario that a designer created.

"A (...) mission is fundamentally just a container for code and data similar in concept (...)

Inputs can be passed into a mission (...) to do things like say inject the type and quantity of cargo whether the ship should be damaged or not where the ship should be heading into a generic transport scenario so that we can reuse the same framework for a lot of different situations. Missions can have content tags (...) to denote their composition. Missions can also have somatic tags applied. These don't specify physical content but rather detail the headlines story behind the scenario that a designer created."

- Tony Z

So designers can differentiate between a straightforward pirate encounter versus one that includes an ambush. Once these descriptors have been assigned, the missions are injected into a database that other systems can access so that they can be instantiated and customized depending upon the situation.

[Problems of the static Mission Design]
This however is where we start to run into some major problems. Designers have to explicitly link missions to various constructs within the game. An NPC mission giver, i.e. or a probability volume covering the area around Yela. This is time-consuming and greatly complicates maintenance if a content tag like 'Pirate' is split into sub tags like 'pirate light' and 'pirate heavy' so that we can have more control over what we're aiming to instantiate - you have to go to every place you design the original mission category and individually determine which of the new ones should be enabled. Worse, this isn't dynamic at all. Mission content is fixed to do specific NPCs and locations so there's no evolution.

The universe is completely static another major problem is that we don't have sufficient context to properly customize the missions at a macro level, the game doesn't really understand the concept of a route, much less what miners traversing that route should be carrying or what condition they should be in or which direction they should be if they're going you know if they're empty heading towards a mine or full bringing the results of their efforts back to a refinery.

This lack of context means that we can't exploit missions ability to have information injected into it for customization and that means that you see a lot less diversity despite requiring just as much work.

The last problem and it's a big one is that there's no NPC regulation of the mission content. Missions are the basic units of work for the economy and if someone doesn't do what needs to be done then everything grinds to a halt. In a properly functioning system, NPCs need the ability to step up, if players aren't going to and the risk award justifies the effort and do the work themselves. If a commodity can be purchased at one location and sold at another for a dramatic markup and there isn't enough cost or risk involved in a transport to warrant the different, the differential NPCs should seize the opportunity and move the material themselves, until the price starts to make sense.

[Economic Nodes]
So let's talk about economic nodes now. An economic node is an entity within the game which might be represented via an actual storefront, a kiosk, a UI interface or an

NPC that possesses an item manifest detailing the items that wants to buy or sell storage capacity and a fair bit of data related to determining prices this includes refineries that process or factories that produce goods and retail stores where you buy finished products factory inputs aren't explicitly denoted but are instead derived from the

//

production formula of whatever they manufacture prices what economic nodes are willing to pay for their inputs and what they want in order to sell are determined algorithmically based upon the rate of change of their of their inventory versus the tangent of our pricing curve. This means that they're smart enough to say raise the price of what they're selling even when their inventory is low if they detect that they're being resupplied at a sufficient rate to get where they want to be in a reasonable amount of time. So clearly there's some dynamism happening here  which is fantastic. Unfortunately, that's where the good news ends. As with missions economic knows don't have any real context in terms of supply and demand and there's no systemic flow of goods throughout the economy why does the refinery want ore ? why is an nPC willing to buy some drugs for you from you where does the shop get the goods that it's selling. Right now the answer to all these questions is the same. we fake it. economic nodes conjure up their own supply and demand out of thin air if a node is designated as wanting to purchase a particular item for example designers dictate a formula that describes how that inventory will be gradually burned off so that they want more. these products you deliver to a node then they don't actually go anywhere, there's no real demand. This is impossible to balance because, while production and consumption are fixed the player count isn't. Sometimes a hundred players are interacting with an economic node and sometimes none... and I mention this because behind the scenes at the economic level all of the players despite being on different servers really are connected to one single system. A real economy is a tangled web of dependencies and you can't expect logical results when it's gears can completely seize up due to the action or

inaction of players.

What we really need then is NPC regulation of the system for NPCs to purchase items when they need them, thus keeping the demand side logical, while stepping in to help with the supply,when players down and the risk reward warrants it. This has ripple effects in two other areas if the demand for missiles spiked because there was a lot of combat. You'd expect that the commodities used to construct those missiles would start to rise in price and you'd also expect to see more miners and freighters working to alleviate that shortage. If the price got high enough that those transports would be carrying a lot of value so part you know you'd expect piracy to increase at which point you should start to see a lot more

security patrols and requests for combat escorts. A simple change in demand men should be able to alter not only what you see as you travel throughout the system but also the available missions these kinds of knock-on effects in fact should happen regardless of whether any players in an area and that's a separate problem I'll talk about more when we geta probability volumes.

So let's take a quick look at what it means to set up an economic node. As I mentioned there are about two hundred and twenty five economic nodes within the stanton system and many of them deal with a lot of different items. Each of these items can specify its own production or consumption formula along with a lot of other information: storage space optimal inventory level price offsets and much more. This information can be archetyped but of course the world's a lot more interesting if every economic node is a bit different so there's still a lot of customization that winds up happening so now. let's take a look at the economic nodes around a typical planetary system. and finally

let's take a look at the economic nodes around the entire stanton solar system. that's a tremendous amount of data to set up and maintain as we're routinely adding new items and changing prices and the worst part is that it's all extremely rigid. There's no way to have an economic node increase how much of --

that the wave functions collapse at which point we instantiate them and start simulating the full-blown entities one of the quirks with probability volumes is that they usually represent a variety of low frequency events space is big after all and even in a pretty heavily populated area shouldn't be running into other ships all that often. the problem is that the likelihood multiple events happening simultaneously is a product of the probabilities, meaning that it's very unlikely you'd ever see two or more at the same time. so let's go ahead and see what this means in practice. here's a player traversing a probability volume he makes it through a good part of it before the freighters probability curve determines that an encounter has occurred at which point the probability volume queries the mission database with the freighter tag and instantiates a compatible mission. the player continues on and shortly thereafter the pirate's probability curve indicates a hit which triggers another lookup into the mission database.

there's nothing fundamentally wrong with this in fact it's pretty accurate representation of what you realistically expect to see two rare events happening on their own schedule the systemic behavior of the freighter and the pirate, even if spawned separately would allow them to logically react to one another so if they were spawned in close enough proximity the pirate might even attack the freighter.

the problem is that this approach lets the algorithms dictates take too much of the experience and if you're not careful that's gonna lead to the gameplay starting to feel very formulaic. what we really want is the ability for designers to craft more custom content and then have a mechanism by which we can trigger it such that it all feels logical if a bit lucky in the timing. so what we do instead of basically allowing the algorithms to entirely control what you see after is we defer the activation for a short period of time and bend the probabilities of other things so that

we're more likely to get a combination that would otherwise be the case.

so they once again trade this the player once again triggers the freighters probability but rather than you already - got it yeah. yeah -- so we're we're running a little bit ahead but basically what you've seen here is the player hit the freighters probability held on to it we bent the probabilities to increase the likelihood of anything else being able to be attached. in this case that that involved the freighter. they didn't involve the pirate rather. so why does this matter think about a pirate and a freighter being created separately. it's going to

take a good while for the pirate to close the distance with the freighter disable it part next to it and board it. this means that you'd never come upon a situation where the freighters already been disabled and the Pirates are rampaging through the ship. so you'd better hurry if you want to save the crew that's actually a really interesting scenario though one that be a lot of fun to experience so while we're fine with the systemic functionality ,driving things by default we want to be able to bend things towards more interesting situations without it ever feeling too random every once in a while.

the problem with probability volumes then all comes down to their construction each one and there are lots of them contains a litany of curves and associated data and is set up by hand which is time consuming and requires constant updating as other things change. one of the main reasons why they're so difficult to maintain is that many of the variables that a probability volume details should impact one another. more security in an area should of course

impact the number of pirates but because they're all handled independently designers have to be very careful when modifying them. mistakes are inevitably made which results in a lot of sleuthing around to figure out what broker is now poorly balanced and how to resolve it. the worst thing though is that they're static. if an area has a lot of pirates it will always have a lot of pirates. there's nothing you or the NPC population can do to change that. rewards won't vary based upon the level of threat in an area because the threat is always the same. a new commodity discovery won't cause miners to move in to exploit it and

won't in turn attract pirates eager to prey on them. demand for security patrols and escorts for that area won't increase in response to the pirates and the pirate threat won't recede as a result. what we really need then is a way to dynamically derive the probability curves based upon what's happening or rather should be happening when we manage to resolve all these problems I've been pointing out across the solar system. So I've called out a lot of very serious problems .things that require enormous amounts of bandwidth to set up and maintain and it still failed to

deliver the experience we want. what we actually want is a dynamic logical universe. we want the demand for goods and services to be the result of a legitimate economic need and we must have sufficient runtime context so that we can customize these modular mission templates. NPCs must be able to contribute in helping to turn the gears of the economy from buying and selling goods to creating and accepting missions. the system needs to automatically balance itself so that we can focus on

the high level rules instead of the details. all of this too has to be incredibly computationally efficient which means we need a more focused simulation engine one that's only concerned with what we need to achieve the desired effect. if we're ever going to reach a point where we can deliver new solar systems in a reasonable amount of time then we're going to need to make a quantum leap.

What's the solution? (...) So this is quantum and I know what you're thinking I've seen this before if the visuals look you know particularly familiar it's because we started with the solar system map that we released years ago but this really is a totally different beast the original Starmap reference static data there was nothing dynamic about it but quantum is totally different for starters it has access to all of the backend services and can pull whatever data it needs from them everything you see on the screen then that's real-time data pulled from the game. Let's go ahead and take a quick tour of the system (...) and jump on over to everyone's favorite gas giant Crusader

And in the background there you can see Port Olisar let's go ahead and bring that out now let's I mention that we're connected to all the real-time data.

let's go ahead and see what the players are up to. Each of those red dots is a real player thousands of them across well over a hundred servers now to be honest we recorded this earlier mainly because we didn't know how many people would be playing the game while we were doing this show right and I figured it would be totally anticlimactic to say here all the players and three dots showed up.

Let's go ahead and continue the tour of the solar system and jump on over to Daymar. And you can see a number of people affixed to the planet some of those obviously you're going to be at an outpost, some of them are inevitably mining. Over there on the right you can see the the current player list. Let's go ahead and see what a couple players up to just go ahead and highlight them so we can see where they're at. Go ahead and jump to surcoworm (sic!) and see where's at.

it looks like he's in Loreville. (...) You can see some deaths down on the planet there, obviously there's been some conflict there and you actually see let's go ahead and back out a little bit and find there's a contract to the right you have access to missions so you can see that Clipsos Norwang (sic!) is requesting personal transport to Loreville and he's willing to pay 4500 dollars. Jump on over to Arial and we can get a view of those probability volumes that I was talking about earlier

What you're seeing here there as I mentioned there are a variety of different types and lots of different content and thematic tags for all these things so you can see everything from the default tags you can see the security reinforcement tags, the bounty hunter tags, the spoofing tags

There are a number of these. We'll be adding more. In general what it's doing is as I mentioned it's basically representing things in probabilistic state, because it's too expensive for us to fully simulate them. (...) Jump over to a shop let's check out the Hicks Research Station, that's on Cellin. (...) Open up that shop look here's yep there's outpost Karea.

And you can see that there's a lot of a lot of death around there and over on the right at the Hicks Research Station you can see all the things that it buys themselves and what those prices have been doing over time. Now I suspect I know what you're thinking at this point: this is all very cool but it doesn't really solve any of the fundamental problems that I mentioned earlier and that's totally true.

Clear the board and get to that so the primary reason for Quantum's existence is to enable us to have one unified world where we can you know simulate millions of NPCs and feedback into that all the player actions into one unified whole so there's really no difference there are some optimized areas the probability volumes that exist in between these two you know these two realities but really

this you know this is this is the full-blown simulation it takes into account all the player actions so let's go ahead and start by bringing a thousand ??? to life so you know Jake here is cranking the number of quanta up to a thousand and they're basically gonna come streaming in from another system oh and into the universe there these are simulated into these and some the real significance of this is when you're talking about working out the vast multitude of details that we

really need to allow this universe to evolve and feel dynamic we don't actually need all of the incredibly high fidelity super computationally expensive stuff that we would get if we were actually simulating these NPCs the traditional way on the server so that they look exactly like they would when you see them these NPCs don't need to do animations they don't need to do physics they don't need to do collision detection they don't need to do a whole slew of things that we couldn't afford to do would be incredibly expensive to try and in the end even if we did it it wouldn't make the end results any more accurate so we also have personality traits - yep one has a little bit of something going on yeah there are a

number of different traits that were modeling we're still working out the exact configuration one of them is ambition and that specifies how far a quantum can push themselves relative to others whether in crime or legitimate enterprise another one is intelligence and that dictates what sort of things they can pursue and how well they can do something happiness is a measure of whether a quantum the singular form of quanta is miserable or content and unhappy ??? want to change their situation move to a

new location change occupations and that sort of thing aggression controls the lens to which a quantum will go to achieve their objective a business-oriented quantum with a lot of aggression might push their workers harder trading their happiness for more profit and as a result have to deal with more turnover and lastly criminality that determines the amount of criminal behavior that a quantum will consider no criminality equates to a law-abiding quantum having a little means that they might be willing to commit some minor crimes like smuggling or illegal mining and a lot means that they're open to piracy and murder they can have possessions store up money own a ship or property have a home now all these ??? have basically come to rest within the system and the reason for that is because there's nothing for them to do there's no reason for them to get

up up off of the couch so let's jump to the next slide we're going to add a little bit of interest to the system by adding an aluminum line to Delamar so we've now injected a mine on to Delamar and yet you see the quanta still refused to do anything and that's because there's no economic reason for them to do it there's nothing to do with the ore that they could go and extract let's go ahead and open up that card though that represents the mine on

Delamar and you can see a couple of stats towards the top quantity and purity and what these are is basically telling you how much of that material is in that mine and purity is to some degree a measure of concentration it effectively details how easy it is to extract it there are a couple of parameters at the bottom that our temporary those will eventually be

derived from those top two parameters but for right now we can override it for

some of this for some of the testing that we've been doing so let's go ahead and jump to the next so now we're going to go ahead and add an aluminum refinery to Crusader and immediately you will see the Qantas start springing to life and

what's happening is they've all figured out they've that refinery it wants to build up a stockpile of aluminum and in order to do that it needs some raw aluminum ore and this it's willing to pay for it and so the quanta or our seizing upon that economic opportunity heading out to Delamar doing the work extracting that ore and taking it back to the refinery and this will continue until it develops you know until it's until it's full so let's go ahead and open that card up again and we can watch the refined aluminum ore gradually build up you can also see the number of workers a few of the quanta so you see many of them actually acting as miners there's also a number of them that are working in thatrefining factory refineries can't process this material you know without labor and there's a formula thatdictates how much labor is required in order to process this stuff and so theyhave to hire a sufficient number of workers and act in order to actuallyexecute this operation and so this would continue for quite a while let's goahead and speed up the simulation a bitand go ahead and keep watching the aluminum inventory is that's finished so they're now building up their raw ore to basically load up and after that the

entire economy comes to a halt they're the one refiner we've added has no more incentive to buy more it's it's it'sstorage houses are already full and therefore it's not willing to buy since it's not willing to buy the quanta have nothing to do so what we're going to do now is at a factory but before we do that let's get head on over to the power plant formula and you can see here that building a power plant just like refining or requires something it requires ten workers working for a duration of sixty ticks and it alsonotably requires two units of refined aluminum or two in order to produce one power plant it's got to be more complicated than that I promise yesagain where we're building upso let's go ahead and add the factory to

Hurston so damn well going too fast there ondown and so once again you seem to see the economy spring to life and nowyou've actually got two different tiersgoing the factory let's go ahead and open up the factory the factory just like the refinery needs workers in order

to build these power plants it alsoneeds refined aluminum so now you've gotmultiple things going on the factory iswilling to buy refined aluminum someonehas to transport that aluminum fromCrusader to Hurston so some of thequanta are basically working you knowthey're River they're basically workingas freighters to move that you know tomove that product from one location toanother some of them were working at thefactory on Hurst and some of them wereworking at the refinery on crusader some

of them are heading back to Delamar toactually do more mining so you've got a entire little you know economic cycle going here however it's gonna wind up coming to a dead end here let's go ahead andaccelerate it again just like what we ??? saw before because eventually the factory fills up all of its warehouses it can't take anymore therefore as nomore don't you know it has no more needfor refined aluminum that wileventually shut down the refinery which will in turn eventually kill off theminers and you can see there that theinventory is still building on thealuminum which is why there's still someactivity but as soon as it's done itstarts flatlining so they see that idea inventories capped out yeah and nowthey're almost done the factory is basically got but it's it's got all it can handle the refinery

has all of its hand to handle therefore the entire loop you know draws too close again so let's add one more node a let's add a shop that sells power plants to our corp slow down again and so now the loop is a bit more complicated the shop is lookingit's a retail shop it's looking to sell power plants in order to do that it needs to be able to buy them for one price sell them for another so it adds its mark upon and it just like the factory that requires transport of the refined aluminum from Crusader to

Hurston it needs the power plants transported from the factory to it and so if you notice there if the individualquanta they'll actually have a little

translucent circle around them if they're carrying goods and so you can easily differentiate whether or not the ones that are moving for example fromHurston back to the mine are empty they're going to bloat up on ore and then when they're heading from or back to Delamar and then from Delamar back to Crusader they've actually all got oreand so this loop will actually continue indefinitely and the reason is becausewe've cheated something for tonightwhich and it sounds very similar to whatI was basically saying was a bad idea

which is the shop right now is a simple consumer it's burn it's burning off that inventory the difference here is thatthe solution to that is very simple inthe context of quantum and the reason ist hese quanta the next step of what we'll be working on is they'll actually require chips chips require engines andso all of a sudden there's a real demandfor how many power plants do you need how many ships do you need it all follows those same illogical equations so let's go ahead and add a bit more economic complexity

so that's far we've shown a very basic economic clue one commodity one refinery one factory in one retail shop the real world of course is far more complicated you'll have lots of different commodities refining and production those retail shops and horseplayers in nquanta competing against one another for a limited supply of goods and services it's this competition this economic natural selection that ensures that things remain in balance that a logical equilibrium is reached so our current

architecture basically has players connecting to a game server and the SIRT and then we've got services behind the scenes that can feed information to these servers and receive information back in return and so you can see here you know we've got shop services and probability volume services and all this

sort of thing the basic problem with this arrangement is that services don't have any knowledge of what NPCs are doing or rather what they should be doing so that shop service that controls the prices it's got algorithms but it has no real understand it doesn't understand that there are 10,000 NPCs that do or don't need this or what pirates are doing or any of that

we don't simulate at that high level anywhere within the game right now you have servers that basically instantiate the stuff near you and you have the services that can run the you know these formula and that's basically it now they can implement these servicesthese algorithmic price calculations but supply and demand only comes from players and hard-coded formulate and it's the same permissions in Peavey's the data is completely static and can't evolve the number of servers of course as we've seen here it varies depending upon how many players are in the game so let's go ahead add quantum quantum changes this in dramatic fashion it

gives us a place where we can efficiently simulate NPC behavior and then feeds that behavior to the game server so that players can experience it further it allows player actions to befed back into the simulation so that there no meaningful difference between the real and simulated worlds quantum creates a complete loop NPCs can be simulated in efficient fashion andactually meant to say that quantum feeds

the services which in turn feed the game servers so let's go ahead and get back to the simulation and what we're going to do now is alter the powerplantformula so let's add a bit of complexity by changing the production formula for power plants to require lera night add some we need to change now we need to

change your production pours I hope right yeah okay yeah we do when I have the new mind first so we're gonna add a lair night mind to wala and then we want

to go ahead and add a refinery for this lair night and we'll add that to the hearse tournelles three and now let's go ahead and adjust that powerplant formula and we're gonna change the powerplant formula right now it requires zero lair night we're gonna change that to two and what we're gonnavery quickly see here is that we'regonna start getting some glare night

flowing to the Hurston factory so that the power plants can continue to be created and that's that's that that says ??? that's that loop of guys right there that's what they're taking now one of the important things here is that the let's go ahead and look at the Hurston aluminum aluminum now what will end up happening there isit was very low for a period of time and that's because they were actually

burning off that aluminum inventory as fast as they could get it and the reason why you see that spike there in the aluminum inventory is because as soon as we alter the formula so that power plants now require lera night they were still receiving from all these Freight freighter transports they still had aluminum coming in but they could no longer produce power plants so it's stalled and all of a sudden they started

stockpiling stockpiling and then when you see that the aluminum started falling that's because finally the they started receiving supplies of lera night and they were able to resume the production process and so all of a sudden their aluminium started getting burned off a natural rate ,,, yep so now we're gonna add a bit more layer and I out in the asteroid fields and Jake's gonna handle that be a macro and now let's go ahead and increase the

supply of aluminum and see what happens so we're gonna add a few new mines so let's add an aluminum mine to sell in

first let's add one to Aberdeen as well and lastly we're gonna do Leiria and if you remember we already have an aluminum mine on Delamar they're telling us we

need to hurry up you guys want us to hurry up or keep going all right so it's like four hundred to one let's add a couple more aluminum refineries as well to the Crusader l4 and an arc or Bell one an arc or pull one yeah so now let's go ahead we just added all of this new aluminum so let's head back to that Hurston Factory and

see what aluminum prices are doing there what's that factory having to pay for aluminum with all this new supply these new mines these new refineries and there you see that the price is starting to fall off you speed it up speed it up a little bit let's let you see what itdoes and if you see it continues to plummet

so those prices are pretty much falling off a cliff now from all that extra supply that's been brought on without us having to go back and rebalance anything so let's go ahead and head on over to Cellin and you can see that Cellin is in very close proximity to Crusader so you've got to refine aluminum refinery and aluminum mining yet there's very little activity and the reason is because the aluminum at that mine is very difficult to extract so let's go ahead and adjust the mining time on Cellin from 180 to 120 and with this again reflects is how difficult is to

extract from that particular location and so you see there now you see a few guys are actually willing to put in the amount of work necessary to extract it not all of them it's still a lot of effort and this comes back to every one of these guys having their own individual set of traits so some of them were willing to injure yeah more risk more repel but you know more effort that sort of thing so let's go ahead now and increase the light consumption let's just go with normal pace all right hey go so let's make the lira night moredifficult to work with and more rare and see what happens there so we're gonna adjust the global Lera night refining time from 60 to 120 adjusting the time on hawala we're gonna do both the global refining time we want to take it to 60 to 120 and wall is the extraction of time goes from 60 to 180 that's right

60 to 120 their night so what we've just done is basically make it more time-consuming more difficult to refine it and we basically made it more difficult to pull it out of the ground so let's go ahead and apply a macro also to Rafal to reduce the default purity of layer night from fifty to sixteen percent which means that refineries are going to need more or to produce the same amount of refined product reactive and now let's go ahead and take a look at the layer night prices on the Hurston factory so we just did three different things to make it considerably you know more expensive for lera night to be utilized and speeded up and you can see that larry night prices are going up up it looks like they're

starting to skyrocket this could be a complete economic disaster if you notice they're just going up up so what we're gonna do now to bring that back in the control is let's go adjust the power plant formula to compensate a bit so let's adjust it from instead of two

layer a night let's change it to point one so we just reduced the amount of ??? ???? needed in the economy by ninety five percent and let's head back and check out the prices again and what we should see here in a minute is ??? ??? should start to drop off pre-significantly there you go crisis averted and this will continue to drop if we wash it you always get it a little bit of turbulence because you've already got

contracts you set up themselves right go ahead and leave it in I just wanna see it drop a little bit more I want to see it Tony so it's already down 40 to 50 percent and it'll keep it'll keep falling okay so now let's go ahead and add some additional factories Delamar looks pretty dead there isn't a lot of economic activity happening over there so let's add a bit of life by adding a new power plant factory and yeah let's add a Hearst nail for factory for power plants as well to create a little bit of competition for aluminum and then let's head on over to Hurston Hurston Factory and let's check out their workers you notice that their workers are falling off a cliff what is happening well we just open up a new power plant factory and they're getting offered better wages for another place and and what are the wages on on on Hurston let's see what the wages that they were offering I can't fit so they were up at that point two-five yeah Delmar's offering like almost three times the weight is yep so there you see the differential which is that factory is more desperate for workers and so you

see let's go actually go let's look at the --- the basically the workers skyrocketed on Delamar and plummeted on Hurston and eventually those wages of course stabilized they find a point of equilibrium which is a big you know point of you know all of this so let's go ahead now and make some major changes we're gonna do this with a macro and you can see where we currently stand on slide there's like a bunch of stuff that we were supposed to be showing here absorb it quickly everyone there we go so we've added lots of aluminum titanium

and degrees iam and the entries that you see on the screen mark system all of the moons now have deposits we've added a bit more layer night but it's still

pretty rare we've also added a lot of new refineries we've added cooler and quantum drive recipes to most factories and they require aluminum layer a night titanium and degree CM the demand for coolers in quantum drives has been introduced and we've also increased the demand for power plants and lastly we've increased you could see there at the

bottom of the screen the number of ??? from 1,000 to 2,000 so that we've got enough workers to keep the economy humming now previously we've seen pretty

obvious cause and effect at this point though the economy is starting to get pretty complicated and we can look at a few graphs just see what's going on with some of the prices to see this and this is one of the most interesting things about a really complicated economy which is these changes in quantities and prices the purposeful movement of ??? this dynamism these are all opportunities that you'll you're going to be able to exploit within the game and therein logical motion so that's it the next one be a shame if somebody came to steal from these people --- -- bring up slide okay got a real slide all right so now we've we've introduced

some some pirates let's go ahead and so you can see that there are four pirate crack and there's the nova rioters lowriders ninetails and dusters pirates have to return to one of their bases in order to refuel rearm let's go ahead if you notice they head to the areas of highest value if you let's go ahead and zoom in there and you can see that those are missions being created by the NPC is no different than players would in other words you see a lot of deaths on a route that's got a lot of value and go ahead and hover over some of those contracts

these are contracts that are being thrown out by the NPCs on that route what's happening is the NPCs are basically being you know some of the freighters are being picked on by the pirates the ship is being destroyed and now that NPC is has ejected and needs transport back to civilization and so they're requesting that and these are missions that would be fed back into the game that you'd see and you could wind

up accepting these let's go ahead and take a look at Hurston the cooler is upon us out this is where they're operating out of you --- what do you wanna hear to check out the Hurston let's look at what happened to the cooler and quantum drives yeah so since we've now got a lot of pirate activity on that route we want to see what's happening to the cost of some of those Goods for a cool half-a-million

UEC you can get yourself your own power plants it's actually normal piracy is really terrible for economics you guys so yeah so let's go ahead now and really see it over at the coolers or it started it like a really reasonable thing yeah that's that's peg it's just it's the scale so basically they were down around a thousand two thousand and now they just absolutely skyrocketed in price because so many of the freighters moving whether it's supplies are moving you

know the finished product from one location to another are now being intercepted and these cuantas I mentioned earlier it's like they have various traits and some of them are very risk-averse and they won't go down routes where you know there's a significant chance of death and so to counteract some of this danger let's go ahead and bring in someone to fight pirates so here comes some security and what you'll notice is that security is drawn to areas of conflict where there is a lot of deaths and such so they're

gonna wind up gravitating towards those areas where there's a lot of piracy and there you can see I'm going over to speed it up a little bit not 200 it's just 200 too fast difficult Oh what's going on and if you notice by the way the police will wind up patrol Ian's area so they're in constant motion whereas the pirates basically are enabling quantum interdiction fields that's why they're basically picking a route on you know they're picking a particular location on the route between you know to between two locations and they're basically you know lying in wait and the security which doesn't know exactly where they are is scanning the area looking for them and so what you notice here is that the pirate situations starting to look a lot better now that Security's shown up and you would see

---

this reflected over in the prices and so what you have here is a perpetual game of cat-and-mouse to where the pirates look for the areas where they can reap the largest rewards with the least amount of risk and security is then drawn to them and as soon as you get ?? enough pushback from the system in terms of security then the Pirates look fo you know more opportunistic you know pastures and again this is very much like what you would expect to see you know you know an illogical functioning system

So let's go ahead and turn on the grid and what you're looking at here is one of the big problems I referred to earlier is how we wind up getting samplings of this you know of how we generate probability volumes and so what this is showing you is we can look at one of these high conflict areas and it's basically tabulating exactly what's in that area so you can see that there are ten total quantum Freyr eleven pirates five security it also shows you the total value being transported you know exactly what's going down that out? let's take a look at a few other locations here you see these security and pirates are fairly even six pirates seven security you know a decent amount

of value six freighters so it's a there's actually an almost equivalent number of freighters pirates and security this is the information that quantum and it's a small piece of thewhole this is what it would wind upfeeding periodically to the probabilityvolume services so that designers nolonger need to go in and say oh wellwe're gonna hard code a certain amountof security a certain amount offreighters and it can never change weship it that's all that's ever going tobe in this case the simulation isrunning it's constantly ebbing andflowing and your actions over on thegame side get factored into this resultjust like any of the quanta in thissimulationso for years we've had just the backend services and the game servers and thefundamental problem was that we didn'thave a single area where the entire gamecould be simulated where NPCs couldrequest goods and services because theyhave a legitimate economic need and could buy with players to provide thosethings we didn't have a way for thisactivity to determine what you see asyou wander around the universe thuseliminating the need to spend enormous sums of effort configuring vastquantities of static data that couldnever deliver the dynamic experience wereally wanted quantum solves all these

issues it also provides us with a lot more context which means that concepts not previously understood like how much risk there is in taking an item from point A to point B can now be easily calculated information like this is vitally important to the price discovery machinery and as I spoke about earlier in a properly functioning economy such things can and will generate a lot of  ripples we also get the information  necessary to properly customize mission content meaning we can deliver a lot more environmental and mission diversity without having to do any more work I've covered a few of the large areas that  quantum will impact but the effects will be felt far wider how many NPCs you see at a landing zone for example and what they're doing can now change over time  manufacturing towns might boom and bust depending upon how much Commerce is passing through them a didn't give in a moment NPCs can grow and evolve separately from their interactions with players into powerful bosses potentially and you'll be able to witness their growth in lots of different ways criminal criminal NPCs with a price on their head can move around the universe just like a player making time to catch them that much more fun we've got a lot more work to do before a quantum is ready for integration with the rest of the game but it's one of the last big steps in setting us up for a completely dynamic and systemic universe.

thanks for listening