Idea for Shodan


#1

Since many people have complained about the new look of Shodan (That it is very Underwhelming for one) and I know it’s only realistic to try to make her more fleshed out and up & current with technology. Just using CGI doesn’t do her Justice. It’s kinda unimaginative. I’d like to pose an idea that could do both, give her a fresh update and still blow away fans of the original memories of her.
Don’t use CGI.
Get an Actor and create makeup FX like Max Headroom. Look at any best of Max Headroom vids and he’s still the best looking AI ever made. He’s so alive and yet he’s really creepy too and seems like such a synthetic living entity. If you could do that with Shodan, you would blow away the whole gaming market. She would shoot to the top of everyone’s attention. Use the same style of Video FX, Glitching, Warped Speech, Cuts, Speed Up/Slow Down, Blurring, and all mixed with the actors use of Expressions and Odds Quips. It could be the creepiest thing ever put in a game. Study all the Max Headroom vids you can dig up and make notes, then find the perfect actor and video FX Guy. Offhand
I would suggest someone like Chris Cunningham who worked with Aphex Twin and did the All is Full of Love, Bjork Video. Someone with his creativity and modern video editing skills. Check out his Rubber Johnny and Come to Daddy work too.


#2

Getting away from bulbous polygons would be a plus. If she could appear to have evolved further, and not just because of the Rachel stage of her branching evolution, it would be nice these many years on with where real-world tech has reached.

Unless Chris Cunningham was a System Shock fan, or saw a personal evolutionary opportunity, and did the work for way less than he could be expected to charge, that’s highly unlikely, but certainly makes the point for a conceptual approach. Even if it was met a quarter of the way, it could be exciting, and creepier than ever on a budget.

It still seems amazing for the time that LGS was able to employ motion capture, hire a company who had the tech, and a couple people for the physical performances, before it was commonplace, especially under that business model and budget.


#3

Well I meant someone LIKE Chris Cunningham rather than personally. I think he’s way too much of Druggy Flake to ever commit to jobs currently. But he made a great example for the direction I’m talking about. Maybe former FX people from Max Headroom are still around and could do it. That actor that plays Max (Matt Frewer) has worked in games and might still have those connections. It would just be so much more mind blowing to do Shodan like that instead of yawn inducing, fake, & Lifeless CGI. If it had to be CGI than maybe try a different way like some very Retro CGI like used in the original Tron movie might look kinda cool.


#4

Swap out everything pertaining specifically to Chris Cunningham with anyone LIKE him, and the points still stand. I’m not being literalist, here, heh.


#5

Y’all seriously need to watch Lawnmower Man, if only for the parts with the CG version of Jeff Fahey’s head:

I get this feeling this is the kind of imagery that, handled well, could be a lot of fun in a modern sequel to an early cyberpunkish game. I’m not sure photorealism would be as effective in this case.


#6

That’s the worst CGI I could imagine. It’s like the old Beyond the Mind’s Eye student collection videos but worse. The Blocky stuff in Tron looked better than that. That would give the game a totally cheap low budget feel.


#7

Ahhh I remember someone brought up Max Headroom before!

Always very interesting to see how changing up the pitch and adding stuttering to a render can make our brains immediately more intimidated by the clear indication that this is merely an imitation of a human.


#8

Which is just what they did to make the original Shodan so freaky. The sound of SS2 is still probably the best audio environment I’ve ever seen in a game. It was so good that they only needed a static picture of her otherwise to make her personality and our imagination of it, complete. Now with the Max Headroom effect they could complete the image both visually and in audio. But I’m sure they haven’t even considered such a wild idea.


#9

I’m really not a fan of the Max Headroom idea. Not that I was anything of a fan of the character back when it was on TV, either.

I think nowadays, we wouldn’t need to be nearly so overt and slapstick as Max Headroom in order to make a character fall uncomfortably/delightfully into the Uncanny Valley. For a start, audiences are far more sophisticated about virtual faces than we were in the 80s.

It also seems to me like a strange conflict between “use a real head, not a virtual one, to make it seem more realistic!” and “put in fakey juddering effects and psychedelic backgrounds that will grind the gears of anyone who knows anything about rendering, to make it seem less realistic”.

Instead, the design of the character was likely far deeper:

  1. What kind of character will players love to hate?
  2. Narratively, what is Shodan’s motivation for displaying itself as a character?

What image is it trying to put forth? Why? What effect does it want that image to raise in those who view it? Unease? Distraction? Empathy? Is it exploiting the human “Women are Wonderful” effect?

Assume that Shodan has access to essentially any psychological text that has been digitized.

Assume that any representation it makes to the player is intended as a psychological effect on multiple levels.

Assume that Shodan has access to very significant processing power, including graphics-processing.

Assume that transmission is digital.

Given those assumptions, would the image that Shodan chose to render have a filter to make it judder like a badly aligned VCR? Only if Shodan deliberately wants to put in those extra microseconds of processing to add judder, perhaps to make you misjudge its capabilities, or to sell some lie such as “I am based on Mars, crackle my connection is poor fizz…”

Even then, Shodan would know better than to do something so obviously fake as juddering: instead, you’d more likely get compression artefacting (blocks of static, discontinuities, etc) to show restricted transmission bandwidth.


#10

First Max Headroom as in the Series, was never Cheesy. The FX wise. And you can’t tell me there’s ever been an AI character that looks as alive and like you’d imagine a true AI that really can show it’s personality better than Max. Tell me if there is, I’d love to see it. Any movie or game. I’ll tell you one, GlaDOS in Portal was amazing, but she had a body so really more of a robot. If they had to portray her on a screen, she wouldn’t have been near as memorable. Max is beyond our CGI capabilities. The CGI characters just don’t have the personality, they can’t twitch their eyes in the right way or something always makes them too static. They can make them pretty lifelike, but not in the way a computer entity would seem. Max is just totally believable as an artificial thinking being trapped in a box. That’s what Shodan is, and only that way of doing it can look as believable I believe. Otherwise give me an example of anything modern that was done as well or better. Trying to denigrate it by simply saying it was simple and used glitching alone to make it seem robotic is such a lazy oversimplification of what was actually done. The makeup was perfect, the look and style of his clothes. The AMAZING actor for the character, his dialogue, even the simple backgrounds were perfect. I loved that there was no interface, it was literally just a half torso in a box, but they did so much with that. And it’s never been done since. At least not as well. So for originalities sake alone it makes it a hotter idea than anything else I’ve seen talked about over the years. Like you said, you weren’t a fan and probably never watched much of the series anyway and understood where it’s genius was. Although it’s great to see the wide range of ways his character was used as well.
The artifacting you suggest is fine, they could do that too. The video glitching is pretty much the same anyway. Nothing seemed Fake about Max and he pretty much fooled everyone in America at the time. Nobody imagined he was a real actor until they saw the real actor. I’m sure that if he came out the first time in 2019, everyone would think he was totally computer generated. At least until it was debunked.


#11

I think the problem is that we experienced a different Max Headroom, and an entirely different 80s.

The Max Headroom I knew was the UK '85 TV movie “20 minutes into the future”, and ‘85’-'87 “Max Headroom Show”, which was an irredeemably awful, massive disappointment, and killed the character for anyone who’d seen it. As a Brit, having lived through the years of the Show, I’m fairly sure a Shodan that aped Max would destroy the game for me, and most others with the same history.

You speak of “the Series”, so probably were watching the '87-'88 TV series, produced in the UK but not broadcast there because by that time the character was dead to us all. I have not seen it, but I get the strong feeling that you are nostalgically remembering something that was actually… well, an '80s TV show. Like Knight Rider, the Dukes of Hazzard, etc, it will not age well, and “beyond our CGI capabilities” feels… well, simply not believable.

I’m willing to be convinced, though: knowing the series (in which, Wikipedia says, he “sparsely appeared”, most frequently as “comic relief”), can you find even one clip that, seen with modern eyes, makes max look in any way good?

This… this is all terrible:

Not cheesy? It epitomizes cheese.


#12

I think i’ve seen every series the movie, and the series. 2 Seasons. And Cheesy is defined as “cheap, contrived, unpleasant, or blatantly inauthentic.”
The effect didn’t look cheap since it fooled everyone at the time, and if it was released right now today it would fool people now too. And it wasn’t contrived or InAuthentic because again it looked exactly like what they wanted to create, a living AI. Of course if it was done today it would look a bit slicker. It couldn’t use video FX since we don’t have video anymore or Standard Definition. But using an actor with Makeup and modern visual effects is what I’m talking about. Of course Shodan wouldn’t APE max headroom. She would APE herself from SS2 which already she did things back then like Max with her Speech and she should sound the same because the Sound in SS2 was Timeless.


#13

I do very strongly want to discourage the devs from using any aspect of Headroom, whether it be the background, prosthetics, humor or scenery-chewing. I’d be fine with the eastereggs/references that we’d expect for any other AI, but seeing it affect Shodan’s design would be a disappointment.

But… Mike Nystul has a great phrase that he strives to live by, and I try to emulate him in this. “Don’t yuck someone’s yum”.

There’s plenty of yum-yucking on this forum already. Max is something from your childhood which you remember with delight and awe and fondness. You want to share that with people, but, like I wrote over in Nostalgia vs Game Design, if we didn’t feel it back then, you’ve got a hard row to hoe to make us feel it now, when we look at that supercut.

Sure, to me, your definition of “cheesy” hit the nail perfectly on the head in every term, and I remember at the time not being disappointed rather than “fooled”: I was eleven and knew exactly what I was looking at, and did not like it. And before remembering Mike’s wise words, I had sat down, ready to argue that my point of view made more sense.

But why?

I’ve made my point for the devs, and that’s good enough :) I wouldn’t enjoy reading someone’s cynical and negative takedown of Time Bandits or The Princess Bride: why should I be negative about what you love?

So I’ll say instead: there were things I liked about him. I loved his backstory. I loved how he got his name. I loved the countless, sadly-unexplored possibilities of this virtual being, caged in a computer in the days before he could have flown free on the internet. There was a huge amount of potential in that character, and in the movie: I do hope it was explored in the series :D


#14

Again you keep reading more into it. You act like I’m saying to make Shodan into Max Headroom. When that’s not it at all no more than I said that Chris Cunningham should be the person they get. I said that effect is under utilized and it’s better than the simple CGI talking heads we’ve already seen a million times and just aren’t believable. They all seem like a video game and that’s all they were. Original SHODAN seemed like more purely because we only HEARD her speak. Our imaginations did the rest. Chris Cunningham showed how modern and amazing mixing real live action with hyper real CGI could be. So did Neil Blomkamp. Anything less is just a cartoon representation and looks old fashioned.


#15

And my point is: that would look far, far worse, far “cheesier”, than a rendered face.

You claim Max is “beyond our CGI capabilities” (you’re wrong, but that’s another argument entirely).

I instead make a counterclaim: there’s no way for someone in cinematic prosthetics to look like a real render of a person: it didn’t fool us in the 80s and most certainly wouldn’t today. Well, I guess arguably they might look like a render of someone in cinematic prosthetics, but that’s not the effect you’d be after.

And you wouldn’t want to do that anyway, since rendering, in combination with facial mocap, gives you all the upsides of prosthetics, with none of the downsides, and a whole lot more upsides since you can take the animation you recorded it and play with it, either through rigging and weightpainting that changes the way the animation affects the face, by scaling movements, by applying kinematic forces, etc.

It could be we’re actually agreeing, and you’re just arguing for full mocap of all lines. The problem with that is that you then need to mocap every line in every language. Any time there’s a changed line of dialog, or a line added in a patch, you need to not just bring in a voice actor and rent a studio again, you need to rent a mocap system and studio again. that’s almost as bad as renting all that plus the makeup artist again, which you’d need for the prosthetics system.

So the alternative is to mocap a bunch of “base expressions”, the main ones for the range of moods the character is feeling throughout their planned lines, and a few custom ones for specific events (like that expression when they found you shot their dog, and such). Then use a library to move the face appropriately for the phonemes of the speech, and maybe another to have the character’s eyes moving to follow points of interest, and so on.

This, it turns out, can work quite well.

Not perfectly. But far better than would be feasible with a guy in makeup.


#16

I haven’t checked in on DeepFake videos in awhile, what a long way we’ve come!

I love the discussions about how to invoke a certain feeling in the player, and SHODAN’s assumed agency over her own appearance and how she’ll present to you. There’s always been a “Cat and Mouse” interaction style between her and the player with System Shock, so it would be interesting to test those expectations and limits, especially for returning players.

There’s something very classic about the way SS2’s SHODAN sounds to me, and a lot of that comes from the audio treatment and how deliberate everything sounds to me. It’s unnerving to decipher her true intentions, especially when her subject matter suddenly darkens or her voice dips in new ways, and you’re left guessing what that means. Showing the player hints at how SHODAN is thinking should always drive more curiosity.


#17

I’ve already noted why System Shock is my favorite game of all time.

One thing I didn’t emphasize nearly enough is the writing and animation and acting and processing of SHODAN, and how that character is increasingly active in the player’s experience of the story of the game. (One of these days I need to play SS with Story turned to 0 and Combat turned up to 4, just to see how much that alters the sense of SHODAN’s presence.)

This is why the thing I’m most excited about in SS3, from the tiny tidbits mentioned so far, is Warren’s observation that SHODAN isn’t insane. Because that gives me hope that her character, and the story of SS3, will demonstrate that she’s capable of reflecting on her two crushing defeats, and reasoning logically – if from incredibly flawed premises – about why she failed, and devising a new plan that builds intelligently on those conclusions.

I don’t think what SHODAN looks like matters much. It’s how her absolutely essential character drives the story of this new game, and how intelligently she connects her new plan to her past plans, that will determine whether this new game is worthy of having “System Shock” in its title or not.

Warren saying “she’s not insane” gives me hope.


#18

Yeah, you’re right, Flatfingers. Looks are about the least important trait for an AI. Heck, even if she’s been downloaded into the brain of a fat guy in a fursuit… it’d all be fine so long as she’s a smart bad guy.

She’ll be kick-ass, so long as she kicks ass.

People love the Die Hard movies because the bad guys had clear, relatable motivation, and a smart plan. If the story were taken from another angle, they’d be the hero of the piece and we’d be rooting for them.

I think this was what was so off-putting about Typhon in the release version of UA: he was neither smart, nor threatening (despite his threats); and his motivation was nonexistent.

It seemed unintuitive and paradoxical, but by scaling his threatening lines way down, they’ve scaled his apparent threat up, to my mind at least: in the later updates, he’s become more of a looming unknown, rather than a comically irritating blowhard.

That’s my worry for Shodan: the line between irritating and threatening was so thin and paradoxical with Typhon, it showed it’s super easy to get wrong in any of many ways.

Like, the best writing and voice acting and character motivation and clever scheming in the world wouldn’t be good enough without also good pacing of delivery: if Shodan delivers an unskippable five minute sermon at the start of each level, we’d never come to love-to-hate her, because she wouldn’t be kick-ass anymore: we’d just mock her. But the same sermon at the end of a level, we might be fine with.

So many things to get right. Game design is hard :(


#19

Not to dismiss your thoughts regarding SHODAN, which are spot-on, but:

Amen to that… but that’s also what makes it So Much Fun. :grin:

I mean, did you see how fast Warren jumped at the chance to help make System Shock 3? Not everything can be a Die Hard, or an Alien, or a Matrix, but when a chance comes along to put the right characters played by the right actors in the right story at the right time, I don’t see how anyone who loves a juicy design challenge says “no” to that.

If I didn’t have 20 years of roots where I am, I’d long since have moved to Austin and slept outside their door until they either let me design for them or had me hauled off. :D