Star Citizen Live: The Bot Whisperer

30 September 202201:02:381,606 linesWatch on YouTube ↗

Transcript — every row opens the video at that moment

00:00:02
hi everybody uh welcome to uh the last episode of Star Citizen live from the void from my home hopefully forever here everything uh I'm your host uh Jared Huckabee uh if you've never seen Star Citizen live before it's where we take about an hour out of the end of our week and we hang out we chat uh we ask some questions we learn about our guests we learn about their work uh and uh this week's guest is is a is a gentleman uh who's very near and dear to my heart probably has no idea how much so but uh mostly I just like saying his name like I just walk around randomly and I just like saying his name uh it's it's AI director Mr Francesco riccucci and
00:00:51
Jessica how you doing hi all good thanks you know it's really great pleasure to be here with you I I appreciate that we'll see if you still feel the same way uh by the end of the hour well you know it's not the first time that we had together so it's like you know I know the feeling another thing yes it's not the first time you've been on the show but it's the first time we're gonna dive deep into what makes rikucci the who the what the where but why why rikuchi why does everything you know very deep questions you know let's see if I can answer yeah for those of us who uh uh who those of you anyway uh who didn't see last week's show uh we are in the citizencon pipeline we are just uh uh eight uh short days away from uh citizencon
00:01:39
2952 on Saturday October 8th uh because of that and the big push towards 318 and PES and a bunch of people on Squadron like they've been follow years like this uh we are we are we are we are relaxing we are putting our feet up and we are we are we're having a couple personal shows a couple shows where we get to know some folks uh if you remember the old meet the dev series from way back in 2015. uh something like that but more longer for him uh if you're big if you have the big questions like when's this come out and when's that come out and and and why isn't this like this well this probably isn't the show for you so just giving you that warning now so you don't complain later instead we're gonna learn all about so let's start things off uh every show
00:02:28
is uh somebody's first uh while we while you've been done this show several times and we're not uh uh there are folks who are no doubt watching right now who have no idea who the heck you are so let's start there uh who are you and what do you do for a Star Citizen so I'm Francesco Kushi as you said I'm the director in calling Beyond games I started in the company in 2015 it was February um so it's quite a while I'm here and I'm basically responsible for all the I Fidelity eye so everything that you see you know how they behave how they what they do uh what should do which systems do we use uh you know on the on the servers it is what I'm responsible for and with
00:03:15
my teams uh right now the item is structuring two three major teams I mean as people cannot see from the monthly reports we are kind of four now we have a little bit no extra help from the vehicle team uh so basically we have the itech team the official team that content team and the Big O AI team that is like you know there are some people in the vehicle team that you know start to do a bit of AI work Tech Team basically is supposed to empower the other teams with technology uh you know to fulfill the features that the projects need uh the feature team you know it's supposed to take this technology and apply to multiple um features that are requested by no
00:04:03
other requirements of the different games uh the Star season and Squadron and then the content team is mostly made by you know um uh designers and animators so what is meant to be is to take all these Technologies features and build a lot of content um so yeah this is like all the people that I either they either report directly to me or you know I support and you know overview and yeah try to help as much as possible and I'm gonna go full screen on this uh what's on the back what's on the the Whiteboard so everybody's already found out this whiteboard because I love riding on the Whiteboard right so it's like there is a little bit of stuff that I used to explain things so we you know we have this this room now and we you know lots
00:04:53
of people come and we talk about stuff you know I like to draw to help myself uh with drawing so even when we do meetings right now I try to move the board next to the camera and then you know I try I draw stuff so there is some usable uh stuff like you know use Channel definition we are trying to simplify a little bit some of the verbs we use in the actions that I can perform in the world so that we can unify a little bit better and and make it a little bit more consistent some of the things that we have through some high level explanation of subsumption data work so subsumption you know uh low level core uh logic works and you know how instant share runtime data static data on the different behaviors and Mission logic and then there is some more um
00:05:43
you know on the right side there are some of the actions of the spaceships you know that we want to do some of the setup of the usables like counters chairs yeah a little bit of everything like you know uh you said subsumption and you mentioned the monthly report so we're uh one last question before we take a step back to your early days here um folks who follow uh the monthly reports and stuff uh it's probably where they get the most uh AI updates uh in general about the state of this the state of that Squadron 42 Ai and uh the persistent Universe AI in foreign what is subscription so
00:06:30
basically Tony yeah that's good so again talk for the next hour uh so when when I basically joined the company right Tony had this speech for this architecture of AI uh where there is a little bit of separation between what we want to do at High Fidelity what we want to do in a simulation State and so on and so on and subsumption is the name he gave to this architecture um what actually turns you know in reality to be is this a combination of stuff or when it says assumption I think it's always like very very broad because it contains both something that you know a regular AI development you would call for example Behavior trees or activities
00:07:18
or behaviors um from the actor perspective the AI agents and then we also include stuff like missions logic so I think we can talk about what is an eye level and architecture level of what it is that I could implementation site right so I think what subsumption wants to do wants to model the world in a way in which it becomes believable and you know realistic how you do that is with a combination of things because in reality when you do game development especially the eyes are very big chunk of the experience that you have but also it's not something you can see in isolation right AI is the output of the work that all the teams do and there is a really large
00:08:06
multi-disciplinary effort into making AI looking good right people sometimes think too much there's a bit of magic there's a behavior running and then that's it but like AI is first of all a lot of animations because animations is the way in which we convey what actually is happening behind the scene the same is voice you know what they say uh reality right like the smartness of people and why we consider ourselves smart is because we can express our thoughts and how do we do that you know with the language with you know uh you know especially with the language that is a construct that we make right to express feelings and ideas and they are in a way needs to model this concept but also make sure that whoever experience that understands what they're doing because
00:08:53
my classic example is like you you might have combat AI that is great and it's blowing behind covers you don't even see it it comes it pops behind you and it kills you the feeling you have is that they are cheating even if they're not cheating right because it's not fun to play against it so I think we always have to balance between you know having the right tools to describe what AI does what the word does you know what they are supposed to do and then making sure that what we describe translate into something that the players can actually perceive as smart but also fun and also understandable all right so subsumption is basically you know the architecture that tries to achieve this one from multiple directions you know and it tries to expose uh the sort of language to multiple departments in a way in
00:09:41
which they can really perform their duties basically uh two follow-ups one uh anytime a player is killed by AI it's absolutely cheating and you should just come out with it you should just be honest about it we have to you know that's that's what I say when I die and I play games right it obviously can see through walls and of course it's not me being bad absolutely like I'm usually really good you know I'm bad when I dive a player in the real world of course but if I get beat by an AI it's so I do things yeah but the reality is also that that sometimes yes the eye cheats right but I think our approach is usually really to never cheat like the cheat that we do might be more stuff like are you really
00:10:30
perceiving the guy that is you know in this suit or do we read that data right stuff that is more like the cheats for me are more things that simplify the CPU usage right and they like they need to your cheat compared to how the real world senses might function but not in terms of what you should know about the world right what you know about the word should always be fair for the players and then of course you know the uh had a second thought but I wanted to somebody said you mentioned the CPU cycles and stuff uh and cheating and whatnot it's one of the hardest elements of balance just combat a obviously there's many kinds of AI uh you know there's not just combat there's also bartenders
00:11:19
lots of bartenders so many bartenders this book so much so many segments so many stories about bartenders but the I lost my train of thought uh but you mentioned the CPU Cycles is up um one of the things that I've seen they're two they're two there are two things I think I always knew but I didn't really I didn't really understand I didn't really know that I knew them instinctively somewhere down playing video games my entire life I had this I knew this but it didn't really come into crystallization until I came here and one of them is how much balance in combat AI is linked to server performance it it is it is it I have
00:12:11
seen like early tests of Siege of wars and whatnot where the AI is just Robert Patrick T1000 Terminators and just going out and murdering everything and then that exact same AI gets put on an evil gets put on evocati gets put on servers and whatnot and then 50 players jump in or 70 players at the time when the test and whatever and the server tick rate comes down and that same AI that was so devastating in you know just playing on a local machine or a local server that's you know six inches away from your gameplay machine uh is now like a Three Stooges routine you know it it it's like pull your head out of your butt it's like what it's it's all the programming and all the effort and all the uh time and the
00:13:00
Artistry of your team and your programmers to to to create these varying levels and then to put them on servers that are under strain that you know need things like server meshing and all a number of other performance improvements uh I don't know I don't really even remember where I was going with the question it was just it was just it was that disparity uh between the work that that you guys can do to to refine and and perfect a thing and then have it all go to hell with a bad server performance was just uh again it's the thing I I think I knew but I didn't really appreciate until I came here yeah I think I think basically I think well our game is pretty unique on some of these things right for different reasons one is I think right now in the last uh weeks there's been a lot of
00:13:48
discussion in the industry about you know game in development you know and you know how things will actually look behind the scenes of a game like I think we have this special position where we actually show a lot of the game that is still in development right because we are still in an alpha stage what that means is that usually in a development right you you pick your battles based on where you want to be not just where you are right so right now for example you have a certain like things that you're going to change is the time right that you have and the resources that you have and you need to try to balance you know which type of task you want to take so let's say we want to have a combat AI you need to have a specific set of features there and a specific set of functionalities for designers for example to balance the game right balancing means how accurate they are I
00:14:38
often did they should you know which type of tactics should they play uh you know where should they be you know and so on and so on now that one as you say it needs to use a specific set of CPU Cycles to do some of this calculation and you need to distribute this calculation because the more you're doing one frame the more the frame costs right you have a very quicker response time from the eye because everything is not in that server but if the frame you know this frame rate goes down then it means you have less you know more latency on how the player can react to that and solid song right now in our multiplayer world of course we have two parts that is that you want to have a very high frame rate on the clients I want to have a frame rate on the server as well because of course the player might see a word that's 60 frames per
00:15:26
second and then the server might run at Three frames per second for different reasons or ten frame per second or 15 frames per second you have a very wide range of CPU usage right on the server so for example CLG Horizon I think one big Focus for us was trying to optimize that usage overall not just on the eye because of course using a lot of but not only right it might impact the elevator that comes might impact you know physics uh calculation anything pretty much on the eyesight right once we have games you know previous company I worked on you know games that I worked on you know we had like a very single play game games usually you have a very strong control on how many characters you want to enable right so maybe you have 2016
00:16:15
20 enemies active at the same time and those use certain things CJ Verizon you know we had peaks of 2000 right and it's like 2000 NPCs are not that simple like 2000 MBC they might fight play is not just like background NPCs or stuff like this right yeah now can you be smarter on that side absolutely is it something we want to support long run that's sometimes the question we need to answer like are we expecting to have one server with 5 000 MPC activity at the same time as you as you mentioned we are working towards our mesh so then the decision comes towards do we want to invest time right now to make sure that we can run 5000 MPC on a server or why are we expecting that we have 1000 NPC on each server at maximum
00:17:03
right I'm actually with a certain you know a lot of servers 500 RPC you know you need to decide those type of budgets are we expecting to have a population manager that is smarter and decides well there should be 20 MPC here and only one there because nobody's around there should they come out of Spawn closet right these are all pieces that are coming together right now but once you don't have them sometimes you either need to say like well except uh unfortunately we are not there yet with a type of experience or no I want to invest you know two months trying to will be smart like one good thing actually that we discovered right was the The Vision calculation right because on a really high performance server you can distribute Vision Vision calculation you know on different ways that you can consider it either a frequency of update that means you want to update somebody
00:17:53
every second if it's close or if you have second if it's close you know in different frequency if they are far but then after or in the amount of ticks you can decide right so we had like something exposed also from you know our core systems I could say like I want to update every 30 updates you know but 30 updates and three frames per second is different than 30 days to 30 frames per second so we switch a lot to frequency but even in that case even frequency means that if you want to do you know an update every two updates every second for 100 NPC that are around you right it's a lot of updates in that frame but that means that if you have already afraid it is low it would be just lower slower right it's balancing uh there yeah yeah you said 2 000 NPCs that that's that's that's spawned throughout the entire
00:18:42
server and stuff uh in the early versions of C in the first versions of Siege of wars and that I saw anything the event alone was spawning seven to eight hundred NPCs like like and that was one of the great things about that event it wasn't just designed to create new content for the players it was designed to push to you know to push on every all the seams of Star Citizen and to see where we can get them that's and uh we got the hundred player servers because of the research that was necessary of pushing on that stuff we got all these other improvements but yeah I think in the end the one that the version that went live it still spawns four 400 or so you know NPCs uh during the course of the entire uh during the course of the entire event which is a massive there's my one I'm sorry
00:19:31
everybody but in the meantime people can also be somewhere else like where they act debate other locations and there are other NPC or other missions right that can still run on the same server so I'm getting at least I'm getting faster at catching them that's good uh uh Sometimes they come out you know all right that's when I get that's why that's why I try to be so even keeled because when I get excited about something that's when the language starts changing that's why everybody's like Derek why are you never hype It's like because that's how I control my language when I get hyped the language disappears too but yeah and and and just to uh to roll back a little bit uh we're not sitting here I wasn't saying you're not saying that that the AI is perfect and the AI is done and as soon as the servers get better everything will be fixed you guys
00:20:18
still have a you guys still have a healthy road ahead of you uh with work to be done this is the current state of AI is a is a shared uh State between the work that your guys still need to do and still need to push still need to develop and then the the you know the performance and the server and stuff coming into a lot you know that will continue to allow you to push and develop that stuff so it's not the results I think yeah for us I think the funny or you know the enjoyable part of our job is also the pretty much you can continue the forever like you know you you might never be happy with the state in which you are right I think like for example perception you know we could work on making it faster for a really long time other things so I think it's iteration right so we try to for example I think I'm really happy already of the stuff that went in Siege of horizon that it means that it's still
00:21:07
like you know it's not as you said it's not a leg consideration done I could see you way far to be done but you know it was a really good and nice uh step towards you know a lot of improvements that we put in and you know it's um I think you know you need to be also proud of the iterations right usually people would see the final game and now they see a lot of these iterations if you look right after the attention don't feel a lot but I think you know it's also great when you see even from the communities and comparison from how it was you know uh 1.5 with a 3A 317 and I think there's been really a lot of of Champs you know of the game yeah uh the uh sorry somebody Chad said every time Jared Sawyer is a Starliner gets pushed
00:21:54
back a year and my brain went no no and I I got distracted ah the other thing the other big surprise uh something you said touched on this for me was how much of AI is animation driven like like how much how much of a believable and authentic AI some an immersive Ai and an AI that doesn't just stand out and look like trash is based in its animations and its ability to Dynamic quickly and dynamically pick between a huge array of animations and then blend them uh as seamlessly as
00:22:42
possible you mentioned this earlier talk to me about talk to me how important is animation in Ai and and do you and Steve Bender just hate each other yet or uh me and Steve are great because I think we also used to work together I mean in crisis we worked together already across the street um and the animation is like you know a fundamental part like again like animation and communication like in my dialogues like those are what really makes AI standing out I mean for me like right there is a lot of example of good AI you know one classic one is of course fear where if you think about how they implement this not even like you know the technology itself doesn't have to be rocket science sometimes to
00:23:30
give you the feeling that those things are very smart right there was a lot of really smart crafting of the environment there it is smart you know set up under these little designers so that is why I was saying like the the eyes is the product of the collaboration of lots of departments and to make it look good and make it look believable I think you know we what we work a lot with Steve for example is that he is the you know the person in charge of all this look so whenever we do in our previous and you know we always work on iteration and white box faces and keyframe animation where they first need to you know what we what we work a lot is try to iterate through the gameplay because one thing is looking good one thing is looking good for the gameplay because when I make a movie right all these so I think sometimes you might create an animation
00:24:18
that looks perfect but anyway you play the player plays in a way which you don't even see it or it's not reacting correctly as you say right it doesn't you know you cannot blend correctly or you cannot transition from one place another reactions is a really important uh you know um scenario where you know you need to have a balance between the amount of Animation you want to create to get the believability but also structuring them in a way in which it's easier to transition from one to another that the categories that you want to play you know make sense for what you're trying to communicate to the player you know and uh you know use different type of techniques uh that we have so I think basically for example the behaviors for my you know take the way I'm you know I grew up my career uh I grew up my skills in my career is that
00:25:09
the behaviors is critical is the central part for the AI to describe the decision flow decision flow then needs to go into making sure that you know the decision flow goes into what any the admission director wants to show right and this is where I think you know we have this I think it always seems of a company when we talk about stuff but really this complexity like you're trying to solve a complex problem complex needs to be somewhere and I think the critical bit is making sure that the complex part are distributed over the ownership of different departments right so for example animation director describes the flow describes the animation State machine that you want to create how to transition from one state into another you know what is allowed to be done and then the decision flow needs to explain one day I was to do without you know padding those two things up too much
00:25:58
because you want to have you know these things flowing nicely and um yeah so me and me and Steve work a lot together you know is is fantastic and misdirect I mean his experiences is absolutely great for us you know even just you know all the things you know we always discuss and we always you know want to push so is this pre-recorded uh that part was but we just switched alive right now so I could answer the question now we're going to switch back to pre-recorded to mix uh so you talked about your career you mentioned your career a lot then let's let's talk about this you could do anything in the world
00:26:46
you know it it's it the world is your oh so you can go and you can do any career in the world why in the world work in video games so for me it's been I could say myself super lacking super laggy because I always dreamed to do this thing and you know I eventually you know been I've been lucky enough to manage to get it of course it's not just luck is a lot of you know uh effort you need to put and a lot of energy blood and you know and tear but you also in life you always need to be you know be like to find yourself in the right position you know in the right moment uh but basically like when when when my dad bought uh Commerce 64 basically right at the first well I think it started a little bit earlier like I was in my cousin place
00:27:34
and he had like a video back that was kind of a feeling uh you know very popular in Italy at that time a little console you know and that is where Harris I still remember noises I was probably four but I still remember nowadays that that feeling of looking at the screen and I mean this job that was moving stuff on the screen I was like wow this is fantastic right and then my cousin got a Commodore 64 I got a camera 64 and you know I started the program and I just started to immediate him pretty much because a lot loved it to do that and yeah this is where I started a program right that was probably seven you know and at that moment there was no real like in Italy especially it was not so much about artificial intelligence per se but you know for me like it was there were people talking
00:28:23
about it right but there was no computer science you know studies it was like electronic engineer so for me I was always saying like I want to study to become you know in the track engineer and then do video games right um whenever I started yeah those little uh magazines that that had basic programs that that you could type yourself I I started working right and I would never worked I started on a vic-20 went into O'Connor 64. so we're right there around the same time absolutely and and like like the earliest photo I have like the earliest photo I have is a baby and then the next photo up is me on a vic-20 you know learning to read and write by just match the symbols match that entirely that's where that's where my dad put me uh he was convinced computers would be the future uh he was right uh I went a
00:29:13
slightly different way but but yeah so all right Connor 64. you've started programming but again things happen junior high high school girls boys whatever lots of distractions out there how'd you get into video games I was I was quite nerd right so it's like you know in general it's like you know yes I did like um my high school was math based high school you know I started to computer science engineering uh in Italy like you have this a little bit of difference between pure computer science that is part of the science floor let's say where you have mathematics physics you know and then you have engineering informatics uh that is like yeah the computer science part of engineering so I did that one because there was more
00:30:01
specialization about artificial intelligence that is what I wanted to do early at the time and I had the opportunity to basically when I started to do my master thesis in Rome and La sabienza um to also a really good mentors uh one is also here working with us as Andrea they taught me a lot about you know coding about C plus plus and in that lab we were doing a lot of Robotics as well and you know with with our professor we did really it was really exciting right and uh you know there was two main you know classes at that time that was also with Andrea it was like you know the assistant of the teacher at that
00:30:49
time and we were doing you know learning and you and Andreas go all the way back to high school yeah absolutely from University okay okay University done yeah it's still impressive but yeah yeah it's quite a lot a long time and I think like you know you know yes uh super strong background a lot of stuff this is why you know I'm super happy to have even more because like he is you know his skills and knowledge about you know yeah not just about you know what we do but even things that we might want to do in the future right and the more research topics as well um but yeah basically we did like a lot of classes about you know learning about you know vision and you know we applied to Robotics and stuff like this and one once I graduated I started applying to lots of games you
00:31:39
know video games company and I was hoping to start with an internship or something is and then I got an offer in Rome first for a company I was being software for satellites um so I started there and you know you get a lot of answers from video games companies I was like okay I mean this is just pretty interesting and I started there and I was C plus plus code so it was something that I still wanted to do and then after my first week there basically got an offer for an interview with EA for an internship here in Germany uh you know a country that I pretty much didn't know anything about you know at that time but my wife that time girlfriend uh she was studying us for you know six months um uh in New York City uh so she was doing like an artist in the Columbia
00:32:26
University and at that time I was like you know it kind of seems like it's a good opportunity it's like you know we're far anyway you know I get those six months as internship in in Germany and then you know we both come back to Italy and I find a game company in Italy and I was work in the economic I was an internship for this game was called battleforge was a RTS game uh based with cards so it was one of the first year games that was based on trying collecting cards but the cars were used to kind of summon enemies in an RTS word and that you would just use it as a RTS uh actually you know real-time strategy game so it was a bit you know double things that you you sum on your end you know your units through cards but then
00:33:15
you know you you actually use them and um and after that I basically moved to to critic I got enough from crydek uh I started to work for the engine licensing so I was you know hired as an engineer for the engine licensing but I think the good thing there was also that the team was not super big so I had to learn a lot of the stuff for the engine in general and I had the opportunity to work with you know a lot of really really really talented people uh hi Kirsten hiring Carson watches the show I know but I don't think Carson knows who I am it could be I've only talked to cars in like four or five times each time each time I get the impression he's wondering who I am and what I had Mark oh you know
00:34:04
Marco is like he's he's one that's been one of my you know major Mentor you know synchronic as well and um Marco knows who I am because he runs he turns around and goes the other direction anytime he sees me coming around the corner so Marco knows who I am uh you know and I had the opportunity to learn a lot of stuff like not just about AI but also you know I I do you know I recognize like that the big uh big chunk of my my knowledge comes also from learning about other systems you know um because I think having the larger overview how things work I think is is really powerful sometimes and you know despite the beginning was I mean I want to do this other thing you know and I need to fix all this other stuff but you know it was it was really powerful in
00:34:50
terms of knowledge right and this is I think it's very classic in the career where you start as a junior or as an intern right sometimes you do like ah but I need to really fix this bag and you don't really see that when you study why do I need to study that product when is this gonna ever be useful right and then I say oh that's where I was gonna be useful right stuff like this so you're in the game industry now you you you you've worked for a couple companies you're you're you found a place a cry attack with with with people you're you're liking and stuff I guess the programmers there's many disciplines for programmers for engineers uh in how do you end up in AI well I always wanted to do AI basically because I I did like I I come my background as well outside the gaming is about
00:35:39
I was playing role-playing games right and for me the acting right the eye is a lot about acting right so it's a lot like well we are in this world action right and I was like oh yeah hey man how are you doing right you know so it's a lot about telling a story and role-playing game is a lot about playing a story or telling a story depends what you know if you're a player or if you are if you're the master you know a dungeon I mean I played a lot of dungeon dragons uh I love like you know a lot of the the fantasy settings you know I did like LARP you know I did like a lot of of nerd stuff uh that I loved it and I still one you know these are my fun this memory you know and I started playing or playing games actually this is the funny connection with Germany it's my first role-playing game I played was uh the dark eye that is a very popular chairman
00:36:28
came because it's a bit simpler that the Indian in Italy was getting a bit more popular at that time and I was probably 9 10 you know I was at elementary school with my best friends and my best friends all said really a nice shared um law for those passion for those things right so I think AI is always being related to me in a way in which for video games is the closest to the tools that you have to tell a story and to make people escape reality into the world that tells the story that they like right and this is why I've always been a fascinated by that I've always been attracted by that so it's um yeah it's it's it's the the way that I have for using my skill to tell stories to people right that make people live in a different world uh uh sorry Ray Tango
00:37:17
TV what are you talking about you just you just you're you're projecting or something I don't know what's what's going on um uh it's interesting uh I'm noting a couple parallels uh between our pads you know starting you know with early basic programming on the c64 it's like this you know going to uh an academically driven uh High School being very being very nerdy that's that's where that's where we diverge because I I you obviously excelled in your studies I went to University and everything I was counterculture rebelled against my academic High School I I literally my my high school transcripts are are failed and Summer's going fail and some are because I refuse to do the
00:38:06
work I was I was grunge and Nirvana and not wanting to do anything I went my school had a terrible terrible sports team and it was all very academically focused and so I didn't want to do the academics because I was just very wanting to be contentious and contrary and I didn't want to do the sports because they sucked and I had asthma so I went and created you know created like a drama club and and did that yeah because in high school I also did a little bit of I mean not forever but I did a bit of theater you know yeah yeah a little bit name one play One play he did in high school no it was they were a place written by some of the teachers there that they were recalling you know
00:38:54
so I was they were not like full uh you know real play so it was a bit of a nerdy strange approach to to theater right but I was I had fun but you know I was more close to the role playing than actually they did it in that case yeah I know so how was your drama class we were we were do we were doing like like Arthur Miller's The Crucible and being you know dark commentary on the Salem witch trials and and and and and just all Dark of the Moon and and and and and and Children of the heart it just really dark melodramatic stuff and then finally the the last the last show before I graduated Annie or for Danny where I was where where I shaved my head for the first
00:39:43
time it was and was uh uh Daddy Warbucks which is so weird to me about my life these days because now there's just more video evidence than anything but there's very little we didn't have my family didn't have camcorders and stuff like that maybe it's cool you know maybe the school had it at that time now theater on camera looks bad like every handheld videotaping of a high school play looks a hundred times worse than the experience actually was now even if the experience was bad it's still going to look a hundred times worse uh when you uh when you when you see it on uh when you see it on a camcorder especially
00:40:31
1994 you know era you know VH vhsc you know you know cassettes you've made it into the video game industry you've found your calling with AI and your ability to tell stories like this you're at you're at you know you're at crytech you're making you're making friends Karsten Marco uh Andreas you know from uni and stuff like this uh talk to me about how you came to cig how did that work out I I think at that time like basically I think there was the search of some new challenge for me right something where I think there were some because I think like after like during crisis three right we had like a
00:41:19
really nice group as well on my colleagues you know I had like were for a high programmers and on crisis three and then after that I think you know I really wanted us to try something something different and I think cig was a really good place because first of all there was they come on technology the base technology that it was used at the time uh but also I think the project is just you know way over the top on almost everything anything right so it's like there's a lot of everything there's uh very you know a search for the the best possible realism that you can achieve and so on and so on uh you know that mission was really large so I think like you know for me it was just a really good place where to continue to to just grow with my skill
00:42:07
right because I think you need to always challenge yourself here I feel like you know you never challenge yourself and after it's always like more space to challenge yourself so I think it's it's a really good place where I always feel very satisfied of how much I can learn and how much I can improve and there is a lot of space where I can improve and get better right so uh so so you're so you're was there an AI team in place when you came here did you have to build it from the ground up what do you recall about those early days so when I joined I think uh uh a lot of the I was thinking care through our middleware at that time that we're using you know when we're using Catera as a middleware right oh wow but then I think we moved into deciding that we wanted to have more in
00:42:54
in-house development right so yes then I build the team pretty much from scratch where uh Maurice and Andrea have been the two first team members you know on the on the tech side and Janine was our producer since pretty much the beginning um and then the team grew up from there so we were like you know we started with three people pretty much and we started to work a lot close with Tony right uh when we started to move a lot more development internally where we decided to really push for subsumption that time right and then me and Tony pretty much like you know started to uh you know crack on implementing all these things where he made a lot of the external tools foreign
00:43:49
for a lot of stuff and I basically pretty much wrote all the runtime code all the level code up you know how the graph gets executed the roads you know runs all the tasks and so and so on and together also with Mars and you know and Andreas you know is with this background he took care of a lot of the ship stuff especially the 3D I mean we have all this this really unique approach to 3D pathfinding and 3D you know uh movement that Andreas you know has worked on uh you know Mary's is you know supporting me on all the navigation stuff since the the you know URI pretty much and I think that has been the foundation of the team and right now I think we are super lucky because I think the people we have on
00:44:37
board that are fantastic you know yeah we have some new Juniors that started a few months ago to start doing like super awesome you know like they're they're really really fantastic so I'm I'm really excited because I think people will see you know more and more improvements because I think how the team has been growing and now the type of culture we are trying to push um that is a culture where everybody helps each other especially and everybody tries to share as much as possible you know the knowledge that she has we still try to keep you know people uh responsible specific specific areas because each other developers loves to have you know their own expertise in you know or grow in a specific way because otherwise you can never be super good at something right but I think all our stuff is very shared you know he pushes for knowledge sharing
00:45:27
but also sharing you know the usage of the system so that everybody the more you use it right the more people feel responsible to to work on on common stuff and um yeah it's you know we we build it from pretty much you know the base uh from zero to right now we are uh you know me Mars Andrea we have two of those programmer and recording that is like Michael Madison we have you know uh Fadi uh that just joined us as our intern tools programmer uh you know we have uh Alex uh we have Danielle in UK as well that just joined a few months ago and it's fantastic we have Jake we have you know Jacob uh what we have time Trio you know that oh fantastic programs I hope I
00:46:15
didn't forget anybody uh I don't think so but you know we have really really great yeah the team's grown since the last time I was I I was able to see him in person it's hard to keep track over the last two and a half years with you know pandemic and work from home and if you don't have a reason to get on a call with somebody yeah you can have no idea they exist uh anymore it's not like you know the old days or you just walk around the office and you know introduce yourself and say hello it's like that you as you started you got the back half of those names I was like I forgot one person is Chris that also China's recently yes so now I didn't forget uh but like oh like it's great to see like how it's hard to find people to grow a team in a very you know in a way that it's not disruptive of course for different reasons not only because like you want
00:47:04
to grow in terms of skill right you want to uh give people enough of your time you know to grow them into their roles uh but also thinking it's to fit you know from personality perspective and I think those guys are fantastic because they fit both ways you know they feed on skills but they feed also personality they are super motivated they're super excited they're bringing so much to the table even in terms of you know new ideas and new you know um the new approaches well we're we're almost out of time we're hitting the end here there's a couple things I I didn't realize you said you started in February 2015. we started almost at the exact same time I was doing some contract work before late 2014 early to 15 uh stuff like this so they don't count my official you know
00:47:54
full-on full-time employment to like April of 2015 but I'd been here for a while so we started around so I'm older than you in this company well technically by HR standards yes I got pay stubs back in 2016. you know I was trying to stay nice too yeah uh the persistent Universe the first version with Alpha 2.0 launched in December of 2015. so we've been here the the entire time versus the universe has existed in the in the in the in the Pu we talked a bit about the the the the the struggle with the server performance and stuff that that that that that hinders this a bit uh obviously you know with with the development of subsumption you know delayed a lot of things we had to
00:48:42
develop the system before we start building AIS on top of it stuff the split Focus between Squadron 42 and pu has to play some uh way uh some factor in this this is my very long-winded way of saying that the AI and the persistent universe is I I don't think it's we're not telling any Tales out of school it's not where anybody wants it to be absolutely all right it's definitely no way I wanted it no nobody would believe us if we tried to sit here and be like oh yeah yeah I know it's it's great you know we're almost there and whatever that's a Sim that's a that's a simple a simple thing but uh you're sitting here you're Elite you're leading this ship you're you're you're
00:49:28
you're you're you're captain of this ship here how do you think it's going yeah forget the party line forget you know you know to take off the cig badge what not you know I have more you know inside than the low level technical thing I think like what you know we spend a lot of time on this one and I think we're spending a lot of time on making sure that some of the things because AI doesn't grow in a linear way I think most of what it does in game development that's true that's true exactly like you know it's a classic 80 you know you know 80 and then 20 takes 80 of the time right the last last 20 but I think we are we are still at the
00:50:17
at the point where I think not not everything we do is actually visible there not everything is yet live right yeah everything is is yet there because you may maybe you have some behaviors that require some setup and the setup is not yet there or maybe there are some things that we're doing for squadron that we are not yet you know in in npu there are lots of those improvements or maybe there are things that are related to the server mesh I think AI is is in a way is this is a little bit of a laggy child of of development because it doesn't come up together until the end in a way right but so I think if you would whatever game you develop right unless unless it's a sequel of something you have a very rough experience doing the development right at least because
00:51:06
of Animation missing because you know system missing and so on and so on we are in the middleware we have like a little bit of Animation music a little bit of system missing you know a bit of optimization missing and and you experience all of this so I think like the way the way I see it right is like for example we still have a lot of behaviors for example in po running that we want to replace and you know we are you know and it depends on the resources you have allocated to to drive that or something else right so I'm personally seeing a lot of improvements right and I'm really not necessary appreciate but I also a little bit said that people can need need still to experience some of the banks and the promise that we have right I I'm not happy about that but I think sometimes you cannot immediately see
00:51:55
why you need to live with something right I think people will be happy anyway when this is out right and I think there is a little bit of of pain that you need to to endure now of course you know there could be discussions I use this too long not too long it should be less but the desire to say with abstraction it's always you know I'm I'm coming from a scientific background for me everything is based on data right if you don't have the data it's really hard to decide that is this a good decision or is a bad decision you need to always balance it but you know I you know for for me of course like you know I would like to have at least everything you know people experiencing you know what at least at the level in which the servers are when when they're fresh and you have a good experience right because I think you know material the players also are
00:52:43
starting to enjoy that type of experience despite you know a lot of the things you know still missing but for example stuff like you know once we release the reinforcement right you see a little bit of things that we are doing that is really simple and I think you know yes the you know uh you have just a regular fixing as any as any development has unfortunately is that you know when you release in in our environment you know you didn't have that six eight months of only bug fixing that usually a development has you know of something that is that is locked because we are really impacted by a lot of the change that happens because they say you know uh persistent streaming you know like I mean and that one in a way another impacts us right and he said you want to fix everything before you release or you know we have more unidarity process this is these are a
00:53:31
larger decision that we have to take so that people can see the progress that we are things we are doing so yeah it's a complicated question you know I think like you know on specific stuff um there is also a lot of coordination between teams and there are things that you know for example a lot like what I'm always joking about is that if people perceive as an ad bank then it's it's our responsibility 100 right to make sure that that one is fixed the reality is that sometimes you know you have dependencies that you need to to to sort out to get some stuff also uh result but it is this is why I think I like AI so much is that because I really like to have also this overview of things that you know and end up being AI you know and yeah I know again I'm the first one that is not
00:54:21
you know yeah it's definitely not at the point where I'm I'm I'm I'm yet proud of what we have there right but I'm proud of what we are doing and I know that what we are doing is the right thing you know the right direction and I hope like you know of course my my goal is to make people experience it as soon as possible right this type of things yeah I I I suffer from catchphrase syndrome I I I have certain sayings I really like you know the only constant in game development being changed probably my favorite my second favorite but I don't use it very often is a is is it's the last 90 you know the last ten percent that takes 90 of the time you said it you said twenty percent that takes eighty percent there's always a variation but it's but it's still a very famous phrase throughout all game
00:55:08
development it's like yeah it's it's that last it's that last part that takes most of the part um you all you also said you also said something about how you know it can get easier when you make sequels and whatnot I I don't I don't want to throw any shade out there or or or or or or or or or give validation to internet rumors because internet rumors are worth every cent you pay for them which is nothing but you know the the the numbers that we're hearing out there for like Grand Theft Auto six you know for the time that's been in the minds it's like it's like I I the specifics are important the message is that this process is hard even when you're doing you know the sixth Mainline iteration of a game which is actually like the ninth or tenth in your entire in your entire franchise and the companies existed for 20 30 years at
00:55:56
this point it's just there's no part of this if you're trying to do something that nobody's ever done before there is no instruction manual there there isn't there is no time estimate that makes any sense um sometimes I think like even with the best intention sometimes you might take a wrong direction right then you need to also not be afraid of accepting that because otherwise you never move forward if you're if you're like there is this uh sentence I think that we got in our package you know welcome of the delphi's page they will say to not be afraid of making mistakes because if you don't make if you are afraid to make mistakes you don't do anything in life right I I think I think if we're being completely honest with everybody and I'm not saying we as in Sage I mean all people who make all video games if we're being honest Mo there's a reason it's called game development and not game construction that's because most of it is mistakes most of it is trying a thing and going
00:56:45
now that didn't work trying a thing going no that didn't work uh you know some of them are big and famous like you know hover mode and stuff like that stuff but most of the ones you never even see well yeah most of the ones are just that's the stuff that's behind we'd have to do a 24-hour news channel in order to cover most of it that is that is a natural part of game development and and it's it's it's we try to show as much of that as that as possible exposed as much of that as possible without undermining confidence in the thing but even a lot of other Industries do that and you just don't see it right I think there is a lot of failures in I don't know drug research right for medical uh you know things and there is a lot of failures in I don't know engineering uh you know like space spaceship engineering research you know it's not that they get it right always the first time look at you know so it's
00:57:34
just like normal and I think for us it's just very visible but it's really cool I think for people that you know that are interested even in that we're just about out of time now uh uh a friend we we I think we've had a a really nice or really honest or really opening conversation it's it's the kinds I I really enjoy um and this place lets me get away with it because I love where I work um but that said give folks some hope here yeah I'm not saying this don't give them a date don't say this is gonna be this or whatever don't do it again with this but but I know you well enough to know that you're hopeful for the future that that you have a positive outlook on things share some of that with us why why should why should we be looking forward to the future of Star Citizen
00:58:25
I think the experience of Star season would be very unique you know not just on the eyesight but on all the systems right I think like it's uh Star Citizen and squirrel for you though in general right there would be a really you know this is why I'm here like because my you know my take on the company and on the project is that those projects will be you know fantastic you know to play and uh not only fantasticity but to read they will give you some experiences that you will not be able to experience somewhere else yeah fantastic isn't necessarily mean perfect Perfection doesn't exist but it means that you you know for a lot of people this would be unique experience you know when you play the game it would be similar for me you know the goal is to have a similar experience that when I was a child I was playing games right
00:59:13
and it was like I I'm in a different place I just feel inside me these type of things and we have a lot of things in in the pipeline already that would be released you know slowly into interview from the work we're doing you know it doesn't have to wait like you know all this you know they're finished but you know we are already in the process too in the next you know updates there will be more and more of the stuff we are working on and hopefully we will do some other shows where we can go more in details but you know it's not that it the good thing here for us is that it's never like uh All or Nothing it's gonna be a very you get this when you get this when you get this one right slowly while we do it so there's a lot of things coming hopefully we'll do each other that's that's the thing for uh uh a friend there's always a nice show that's good there's there's there's there's
01:00:00
just there's always the week and then the week after and the week after and we don't fit you in this week we'll get to next week or the week after that of the week after that uh it's it's this is definitely a marathon not a Sprint and stuff like that all right thank you so much for taking the time out of the end of your week here to to chat with us to hang out with us um folks watching at home uh this is the uh uh this is the end of the quarter it's the end of our our quarter three for for for Star Citizen the the the the summer quarter for our video content uh we are uh we're we're taking our little break now uh we've got citizencon coming next week which is the on October 8th at 8 A.M Pacific 3 P.M UTC that's gonna that's gonna kick off our new quarter
01:00:50
four uh of content and uh the journey to 4.0 and a whole host of other things you're gonna have to tune in uh to find out uh remember it's a it's a smaller show this year there's no big keynote demo or anything with the focus on 318 and getting persistent entity streaming is this new foundation for everything in the persistent Universe plus moving to offices to Major offices at the same time uh the responsible thing was to scope down uh citizen con a bit this year so so remember that and uh when setting your expectations properly that said that said you know I am not one prone to Hype there's some there's some of the stuff I'm most I in my eight years here there are some of
01:01:37
the most exciting developments that have come to the prison Universe yet in next week's show so it's it's not going to be one you want to miss check it out citizencon October 8th 8 A.M Pacific 3 P.M UTC uh if I'm saying it you can take it to the bank check it out uh yeah that's it I forgot how to end the show I got so caught up in the citizen gone stuff I forgot how to end the show what do I usually say I don't have enough we'll just we'll just here we'll just we'll just we're just gonna hit the thing and just turn it the boy we'll just turn it to Black just pretend that the show ended and that I said something really slick uh attend the show that was really cool yeah what did we learn this week that's for the other show I don't have I don't have a sign off for this one
01:02:25
maybe I need one wow

Transcript from YouTube’s automatic captions — punctuation and Star Citizen jargon are approximate. The video itself is hosted by CIG on the official channel.