shallowgravy

Catch Phrase!

[molinaesque]

Cooper's first reaction to Lucy "At first glance, your treatment of this man appears unfair, and I'm obligated to intervene." "Huh."

[shallowgravy]

I really like his costume also. His shirt is just a dirtier version of his blue and gold western shirt he had during his films. Also, Cooper is wearing a light colored hat, dingy though it may be, wearing a white hat signifies that he's a morally good cowboy.

[toastbutteregg]

[shallowgravy]

True of most games, but Commander Shepard immediately sprung to mind.

[medallionaesthetics-deactivated]

More Detroit content, just as promised. Can you tell who this one is? . . . . . . . . . . #dbh #dbhfanart #dbhmarkus #dbhconnor #dechartgames #dbhkara #dbhart #dbhralph #detroitbecomehuman #detroitbecomehumanfanart #detroitbecomehumanart #lineart #blackpoint #linework #micron #micronart #micronartist #ink #inking #inkart #inkartwork #inkartist #quanticdream #quanticdreamgames https://www.instagram.com/p/CV0yH1KpaKV/?utm_medium=tumblr

[shallowgravy]

That's some nice line work right there. 👍

[nerd-mom]

[shallowgravy]

God, this game is pretty. Wish I could play it for the first time all over again.

[deviated-from-adulthood]

Im not very good with photoshop but I dont see a lot of DBH icons so i was like, ok imma make some.

Credits if the used pictures goes fully to dear op @doesitsparkjoytho thank you so much for your awesome pics. I dont know what Id do without them.

I couldn't choose between filters again so yup XD

Feel free to use any of them and ask for any character that you want icons of. Im not good at it but ill try my best :D

Her life was more important than mine

[shallowgravy]

She ❤️

[deviated-from-adulthood]

a theory of how deviancy works in DBH based on my slight knowledge of AI

ok so its a longgg post. Im learning programing in university and we have a course regarding AI and basic information about it. after my today's class something got my attention and with a lot of research (hopefully accurate) I came up with a theory. its not perfect and can be flawed as Im not an expert in the field so I would appreciate any criticism.

so before I began they are some terms that you need to know. they arent difficult and I tried to simplify them as much as possible but if you have any questions about them or the whole essay in general Id love to answer.

algorithm: is a set of instructions for solving a problem or accomplishing a task. algorithms can be simple and liner or difficult as hell.

a simple example of algorithm: I want to make tea, how should I do it?

I go to the kitchen--->put the kettle on---> take the tea bag--->place it in the cup--->water is boiled---> I pour the water in the cup---->tea is ready.

easy ha? its a liner algorithm. now if i didnt have tea bag i had to return to the first place so it will be like

do we have tea bags -->yes--->the rest

or if the answer is no-->find tea bag

or waiting for the water to boil--->check the water-->water is boiling--->make tea

or alternatively water isnt boiling--->return to waiting for the water to boil

and so on.

data and information:

data is the raw Knowledge that AI gathers and information is what it can process and gather from it.

machine learning: a process in which our AI learns new abilities to perform. and AI learns the new abilities without the need of programing the codes. its when I no longer need to program every single algorithm for my AI. It can write its own by learning and gathering data through practice and training. really cool huh?

target functions:

target function is the result we expect from our AI by giving certain data to it. for example we want our AI to identify red color. we define what red color is (they are different methods for this, we'll get there) and expect it to recognise all red colors in the future. our expectation is target function.

parameters: the boundaries and rules in which AI compares the data with it. for example if you set X=5, your AI always places 5 in where X is in its new given data.

bias and varience:

when you program an AI, you can decide how much you want it to be sensetive (not emotionaly ofc) towards outside factors. by giving the AI new data youre teaching it to use the data for writing algorithms. but you can decide how much information is relevant to the job of AI.

bias refers to the amount that a models prediction differs from the target value. we give data to our AI and expect it to give us answers according to the data. for example if teach my AI to recognize all the red colors, the AI scans the color and decides if the given data matches with its own algorithms or not.

so bias error happens when we make our AI unreliable by too much simplifying the assumptions for faster results. If my AI is given a set of colors from yellow to red and is told theyre all red, it will recognize yellow and orange as red as well. so high bias isnt good for our AIs accuracy. the algorithms will be parametric. same results no matter how much we feed data to it for improvement. you can show hundreds of red colored objects to it and our AI still considers yellow as red too. so the AI is underfiting

variance: variance indicates how much the estimate of target function will alter with different training data in usags. so in other words, it means how much my program is flexable to new added data.

if my AIs variance is high enough, it adds more data in its algorithm creating. a low varience AI that is designed to find the color red will only consider the few red shades it has seen and not many other shades of red.

high variance might seem good, but is it really? Imagine an AI that is created to find songs by listening to them. if its variance is low, the AI will only listen to the closest sounds it can receive and only searches websites related to music. now if our AIs variance is too high, it will analyse unrelated frequencies from far away and searches all the unrelated websites too. so the AI is overfitting for the task.

too much unwanted information disturbs our AIs function from its goal. the AI will take into account information that isnt needed at all. aka "unwanted and unrelated" factors to its mission.

sounds familiar huh?

heres a demonstration of the terms. soruce

now, I have a sub theory of my own that I need to talk about it more in the future but the gist of it, IMO, both Markus and Connors AIs are trained with reinforsment learning. what is reinforcement learning? its a machine learning supervised method based on rewards and punishments. rewarding desired behaviours and punishing the undesired ones.

so our agent (the AI) does an action which puts it in a situation that the superviser can decide whether is wanted or unwanted. by determining the state of action, our AI learns if the action is correct or incorrect depending on our response with award or punishment. for example when our AI sees a yellow color and tells us its red, by our punishment it learns that yellow isnt red and wont ever consider it as one.

ok so hear me out, why would Kamski give his precious prototype to Carl. unless if hes someone that kamski knows can train his android in the desired way. Carl is the perfect teacher in his opinion and we can see how he judges Markus actions and tries to navigate his ways.

what about Connor you would ask? ok so, let me introduce you to our beloved Amanda. who desighned her? yes Kamski.

"but he wants the androids to be free, why would he design Amanda like that"

well first of all, as Amanda herself has mentioned, Kamski designed the base of Connors Zen garden. his code has been tempered with a lot ever since.

and secondly, Amandas role is only a handler. a negetive or possitive feedback to Connors actions. but the decision of which action gets which can be changed really easily. just a change in defining particular actions as bad or good.

so instead of X=reward and Y=punishment we have X=punishment and Y= reward. you can change and replace anything you want. if AI is designed to find red colors and I want to find blue now, I can change the factors that awards red colors and make them to award blue ones.

there is another concept in reinforcement learning; selecting domain. basically the AI can prioritise which handler its going to listen to if they are more than one. the concept of value is also has a part in there. what is value?

value is the long term award. for example in eating fast food, its taste is award and in eating healthy food,  healthy long life is the value. the AI can decide each handlers value and choose who its going to please.

the RK series AI needs to be flexible for constantly changing and reshaping parameters in reinforcement learning with new information. unlike other androids who could only extend their algorithms in the defined boundries set prior to their learning. their input is already set and labeled. imagine youre baking dinner. in first scenario youre given a set of recipes and some ingredients. you bake what you can do with the limited materals and recipes. you cant bake the ones that require more ingredients than what you already have. now in the other scenario youre given a whole supermarket of ingredients to knock yourself out. you can make whatever you want. this makes the food more varied but also increases the chance of messing up too.

so even if we ignore the reinforcement learning method, there are other factors in the game that indicate their flexibility and high variance.

Connor was designed to infiltrate. he can constantly change his approach depending on outside factors (I need information to determine the best approach) he can alter his behaviour in a snap. he can negotiate, interrogate, befriend with his partner and choose his words depending on the information extracted from them.(the ideal partner shenanigans). that scene in waiting for Hank chapter is really goofy at first glance but when you think about it, hes talking about music and dogs cause he knows, he knows it will work on Hank from what he had gathered. he must take a lot of information into account for reconstructions and preconstructions. another android who can do it? Yes, Makrus.

they also have more free will. and their parameters is more adjustable. Amanda cant control Connors actions and Markus can refuse carls demands even prior to deviancy (when he asks markus in Broken chapter to take him too you can leave Carl and go alone)

so in calculation, the parameters can change with new data depending on its flexibility. the more free, the easier to change with added information. a normal android has high bias regarding its decisions revolving around unwanted subiects. for example no matter how much they are beaten or insulted (the new added data) they wont react to humans attacks(the parameter of not hurting humans or carring about damage from them)

how about deviants you ask? we'll get to there too.

so we understood unwanted information can be gathered by our AI if its variance is high. how is all of this information relevent?

lets take in one of Connors missions.

his sole purpose is to catch deviants no matter what. but you are given the choice to let them go several times even prior to deviancy. why? cause you can take in extra information into account. into choosing the next action. a normal android with that mission would shoot them without hesistence, catch them as to fulfull his highest priority. the only relevent information it would consider. yet, he doesnt. high variance, everyone. (its arguable that because he does it to get Hanks favour for the ideal partner stuff, even though that decision, is itself is not part of his program. he is told several times by Amanda that his mission matters more than Hank. she actually gets mad at you if you talk about Hank and your relationships growing. but even considering that possibility, youre still given a choice to shoot or spare North or markus after remaining a machine. theres no Hank there)

emotions are not part of their program or a virus. but rather a learnt behaviour though to their high variance. they must not care, yet they do. cause the unwanted processed data manifests in form of emotion. Im not saying their emotions are fake. they are real because they are not mimicking it. they are acting upon their uncontrollable reactions regarding certain triggers. just like emotions in humans. certain triggers reactivates chemicals related to emotions. causing us to act like that towards that trigger.

How is it that they can make other androids deviants too? they dont give them a virus. they change their codes. their boundaries, they shift their parameters for them. making it more variance so the other androids could see other possibilities rather than just following orders too.

"you say Connor and Markus were always deviant; but they break a barrier to become deviant, how about that?"

well yes, because they are breaking the false walls of parameters that stops them from reacting or disobeying. why? Cause they are learning to do that! they wouldn't do it if the parameters werent flexible enough. especially considering both didnt endure lots of trauma or emotional shock. Markus was just being insulted and pushed around by Leo. and Connor was basicaly told to deviate. they decide to deviate cause they learn from the new gathered information.

maybe Im grasping at straws here but Makrus is even shown to do the calculation. when hes being pushed and decided to deviate you can see how hes taking in extra information. "this isnt fair, dont defend myself?" he's questioning! comparing the event with new data.

so about the other androids.

once again its totally speculation; but in compiling and cumpiling data, if theres a disturbance in the process, an error can change the reading of codes sequence. connecting lines that arent connected in the first place or are locked. so maybe an emotional shock or physical attack can interrupt the execution of programs and accidentally cause the break or alteration of boundries and limitations in parameters. making the androids realize about their situation and existence of other options too. (not...fair/ I coudnt take it anymore and so on) so with the slightest realization they act upon it and become fully deviants

Kara deviates cause her parameters are changed by prioritising Alice's life upon everything. she is programed to care for little kids and seeing one being beaten up is agaisnt her programs to the point that it actually causes an error.

daniel is a family android and when he realizes hes about to be replaced, the part of his program that instructs attachment to family is contradicted enough to make his OS go nuts.

Shaolin, Ortizes android is beaten to the point of error, tracies were beaten to the point of error. north is beaten to the point of error.

they all had objections against their normal way of computing.

oh btw, Connor can choose to remain a machine, unlike others. and why would he choose that? cause in that version, he prioritizes Amandas rewards and his mission over everything. he "chooses" to stay a machine.

what an irony huh?

a side note,  I think the software instability is a form of varience. the higher it goes, the more flexible the program is. therefore higher varience. and when you keep it low enough that you cant unlock the deviancy option, is because Connor is changing his parameters with new decisions. making it less and less unstable. to the point where his software is limited enough that deviancy is no longer an option.

[shallowgravy]

hi, i just wanted to pop in and say i like how insightful you are about dbh! you bring up really interesting points and bring up cool symbolism that never even crossed my mind 👍

Thanks. I'm a few years too late to the fandom, but it's still interesting to think about and discuss. The game isn't perfect, but it certainly is compelling.

[anwynb2003]

GUYS GUYS GUYS-

So, y’know Connor’s tick of adjusting his tie all the time?

Like he does it so often.

And then near the end of the game, if he deviates and brings the thousands of new androids to Markus:

Tie be gone.

He didn’t NEED to get rid of it, like at the beginning when he made it into a tourniquet. He just decided to.

And at the very end when he sees Hank:

Still gone.

Conclusion:

CONNOR DOESN’T LIKE TIES. HIS FIRST SIGN OF DEVIANCY WAS CONSTANTLY FIDDLING WITH HIS OWN.

Idk if anyone’s said this before and I’m way late, this is just something I’ve always noticed. Is it common knowledge?

[shallowgravy]

Yup. Does not like the tie. It might even be symbolic. Dogs are a motif in Connor's story (he is loyal, but to who?) and the tie is a leash he's constantly fussing with, especially when he's a machine. RK900 has a stiff, restrictive collar to symbolize his total obedience.

The game is chock full symbolism, some more on the nose than others. Connor flipping a coin (will he be good or bad). Amanda pruning roses (removing a symbol of love). Marcus activating the birds (he's trapped in a gilded cage). Marcus with either a halo or horns in certain shots (take a guess). Markus has the motif of fire (provides warmth and light, but can also be destructive). Kamski's portrait of himself in a garden with a male and female android statue on either side (God in the Garden of Eden with Adam and Eve, major God complex). Kara and Alice are threatened with a knife and forced to "eat" with Ralph (repeating a cycle of abuse as they were attacked by Todd during dinner just last night).

[catchymemes]

[vincentspork]

who’s putting washing machines in their kitchen

[fairy-islearchive-deactivated20]

British people, apparently

[reematheroamer]

tag with where you live and where your washing machines reside

[shallowgravy]

American here. The washer and dryer is typically in a laundry room next to the garage. However, I remember an apartment in Florida that had a stacked washer/dryer stuffed in a closet next to the kitchen. Placement of appliances may depend on the amount of space and whether you live in a house or apartment, and the apartment complex may have laundromat instead.

My grandparents had a house with a basement. The washer/dryer was in the basement with the bathroom overhead. Even had a laundry chute in the bathroom so you could dump your clothes right into a basket next to the washer. It was awesome.

[angelshizuka]

One of the things I hate most about the “Alice is an android“ plot twist, is how they tried to use Alice being emotionless/down and not wanting to eat as “foreshadowing“ when those are common symptomes of depression/trauma.

I didn’t see “android foreshadowing“, I saw a traumatized little girl who wanted to love her father, but had to learn the hard way he didn’t love her and had trouble wrapping her head around the situation they were in.

And I still stand by that an android rescuing a human girl from an abusive home and becoming her new mom, is at least a 100 times more interesting than *GASP* child androids exists!? (Seriously, the game made that a much bigger deal than it was…)

[shallowgravy]

I didn't think the plot twist worked either. Players who pay attention can figure it out around the midpoint of the story if they open the secrets box and see a magazine about child androids, not to mention that Alice and Luther can strongly hint at Alice's secret if you talk to them. Connor can also outright say "deviants" crossed the highway. I feel it would have been more interesting if the player knew early, or almost immediately. It could add more tension to the Alice/Kara relationship. Alice knows she has to pretend to be something she's not. She believes her relationship with Todd failed because she's not human. She's sad that she's different (the bedtime story can hint at this). She believes Kara won't love and protect her anymore once Kara accepts the truth (actually possible, Kara can abandon her). Kara seems to think being an android is lesser (a dialogue option in the motel has her say "I know I'm just a machine..."). I think her self-deception is related to this. She doesn't believe an android could truly love her like a human would. I feel the narrative would be stronger if Kara struggled more with her own identity. Learn to accept who she is and work through her self-hatred. And if she failed to do so, her bond with Alice would deteriorate beyond repair once the player reaches the Crossroads chapter. Have the option to accept Alice be locked if she failed to accept herself. Alice then leaves you for Luther/or Jericho. She no longer needs Kara to survive (or so she thinks, she will end up in the camp and only attacking the camp would save her). I don't know. It's what I would suggest to Cage. Plot twists can be hard to pull off and he didn't stick the landing. I don't think that's from lack of trying though. The best twists are tied to a theme or a philosophical question. Let's say the the theme is "learn to love yourself before you love others". (Is this what Cage was going for when he made Alice an android? Not sure.) If Kara loves herself, she will find a family to love her and Alice will also love herself thanks to Kara's acceptance and undying loyalty. Their love is healing and they find some peace regardless if they reach Canada. If Kara doesn't, then their turmoil continues and Alice will probably die in the camp. Just a thought about how the "Alice is also an android" thing could work. I'm not a professional fiction writer. It's just fun to think about.

[nock-and-bolt]

Do Converted Androids Differ From Self-Deviated Androids?

I’ve been thinking a bit about this. Mainly how the androids Markus converts have crazy, sudden, and fervent devotion to him. 

The Freedom March?

The one where they assemble hundreds of androids, a lot of them freshly deviated?

Yeah, so remember how you can tell them to just stand their ground, and they’ll just do it? Even as they’re all gunned down? Or order them to attack? And they’ll run straight into a massacre?

I don’t know about you, but I feel like at least some of those guys who just woke up would probably be like…okay cool thx bro for giving me free will and all. But with this nice fresh appreciation of not wanting to die, I think ima skip out on this one. 

But that doesn’t happen. Not for a single android.

AND THEN ON THE OTHER SIDE OF THE DEVIANT QUESTION you’ve got people like Ralph, who self-deviate and just want to stay out of the way and live in peace. Yes, Ralph can sacrifice himself for Kara and co. at the camp, but only if they’d built some sort of relationship first.

Likewise, North and Simon will both sacrifice themselves to save Markus. But only if they’re at a certain level of closeness/relationship. And guess what? They also self-deviated. You know who sacrifices himself for Markus without any relational status involved? John. How did he deviate again? Oh, yeah. Markus pulled him behind a crate and converted him.

And this instant devotion doesn’t seem to be limited to just Markus.

Remember this? Free a bunch of androids from CyberLife Tower and they’ll just march forth all over Detroit at Connor’s behest? And suddenly people want to hear a speech from him? Markus is dead and you awakened us so of course you should talk to everyone and be our leader! 

To be fair, the anomaly with sudden, enthusiastic devotion of deviated androids is likely just due to gameplay mechanics and what’s easiest for streamlining the story. There’s also something to be said for being a bit lost and looking for direction after gaining free will/freedom so suddenly. A leader calling you to do something, offering purpose and direction, certainly would have its appeal.

BUT that doesn’t account for all of it. Especially not with what Jericho is willing to do for Markus. How they follow him blindly regardless if he wants to talk and make friends with the humans or if he wants to wage all-out war against them (looking at you, nuclear ending). I can accept that there would be androids in both of those camps of thought. But for the same group of androids to swing so wildly based on the word of one man? 

Wasn’t it free will they were supposed to have gained? 

And as a side note—

Can we just talk about the difference between when Markus saves these guys versus when Connor does? 😂

HINT: that first pic was when Markus saves them. Second is Connor.

And yes, I realize that Markus is their known leader. Of course they’d be happy to be saved by “Markus himself.” But dayum, son, in such a high stakes, life-or-death situation, I’d think you’d just be happy and relieved to be saved at all. Me personally, I’d probably be ready to thank whoever it was regardless cause holy RA9 I’m still alive.

I could see an argument where, well, maybe he recognized the “famous Deviant Hunter” and was confused. But if that was the case, I think his reaction might’ve been a little more pronounced. More of a recoil in shock/repulsion rather than a “oh. Look. A person. Not my deviator.”

We know that Kamski wanted to the revolution to succeed. There’s footage of him that didn’t fully make it past production where, if the revolution had failed, he says he’d “be at the helm” of the next one to make sure they’d win (timestamp 6:26).

And we’ve seen the variable behavior of deviants who deviate on their own versus those who are converted.

Is it that far fetched Kamski might’ve…added a little something extra in the coding there when one of the RK-model androids deviate someone? 

Just some food for thought. ;)

[shallowgravy]

Yeah, that always bothered me too. Markus' connection to Kamski was the most intriguing part of his back story and the game does nothing with it. Connor never gets to ask Kamski or Markus about it either despite having an inkling about their connection. And no one, including Markus, questions how he can convert andriods (though North seems blown away when she first sees him do it). It seemed like an interesting way to explore what freewill actually means and a potential issue for Markus' cause should humans conclude androids are being reprogrammed instead of achieving real consciousness. I'd say it was rich fodder for a sequel, but Quantic Dream doesn't do sequels. It's probably a dropped plot point like the supernatural elements in Heavy Rain.

[larygosomens]

Crowley just walking in his flat appreciation.

[bonus]

[gingerhaole]

Watching him saunter makes my stress melt. Opposite effect on his ferns, but I AIN’T THEM, so

[shallowgravy]

Wait, is the "wrestling angels" statue in the same room as Crowley's indoor garden? I think it is. Oh, Crowley, you gay disaster...

[crowleyraejepsen-is-retiring-de]

aziraphale’s a shitty dresser and he owns it. crowley knows deep down that he’s a shitty dresser but he’s a coward and refuses to live his tacky queer truth

[shallowgravy]

In my head canon, Crowley will wear whatever is "cool" and trendy. He tries really, really hard to be cool and "with it", which means updating his wardrobe all the time. Meanwhile, Aziraphale wears the same coat every day for a hundred years. He sticks to his favorites until they draw too much attention. He also wouldn't wear anything "provocative" in public, although there was that one Halloween...

[ileolai]

I see ‘’Aziraphale must have blessed the holy water himself’’ floating around a lot but..  the script implies he stole it from the jug, yo

and… i also assume Crowley thinking the font should have guards means he hasn’t seen holy water since Heaven, where this thing must be guarded 

so like, what kind of Mission Impossible bullshit did you pull, Aziraphale?

[shallowgravy]

I assumed he stole it or received it by doing buracratic ninjutsu/lying his ass off. There also seems to be the implication that the holy water is the very best quality. The standard human-made holy water isn't very good, let alone capable of killing demons instantly. Aziraphale wouldn't want Crowley to suffer a long, painful death. Getting the holiest water may very well spare him that, which probably makes Hell's method of execution kinder than Heaven's.

[cliopadra]

So like, does Gabe or one of the other Archangels get directly notified whenever Aziraphale does a miracle on Earth or…? 

[boyfriendinthedarkglasses]

Uriel: “Don’t think your boyfriend in the dark glasses will get you special treatment in hell…”

Aziraphale: #omg wait you think crowley’s my boyfriend??? #omg what noooo no way haha #you really think so??? #i mean like he’s definitely not #but that’s the impression you got???? #not that it matters #but that’s what you truly thought when you saw us together?????

Uriel: “He’s in trouble too.”

Aziraphale: #… what

[shallowgravy]

That first picture isn't *tee hee i have boyfriend*. He's emotionally overwhelmed. That smile (almost a laugh) is a defense mechanism brought on by fear. He's scrambling for a way to react, and possibly deny the accusation, after having his most dangerous secret revealed. Should I ever be cornered by scary people accusing me of treasonous behavior and they were ready to hurt and murder me, I doubt I would be able to instantly respond in a logical manner. 

The second picture is him hearing that his worst nightmare is going to finally come true.

[assiraphales]

aziraphale’s bookshop is really old right? what if it’s haunted but he never noticed

[shallowgravy]

I like to imagine the shop is somewhat alive like the Bentley because Aziraphale loves it so much and he's owned it so long. The more books it has, the more powerful it gets. Aziraphale has added strange, unearthly geometries and pocket dimensions just to store more books. The shop will straight up attack shoplifters and will "glare" at customers until they leave. Crowley is the only demonic influence the shop lowers its defenses for because it lowkey loves him a lot. He can even touch and read the books if he wants...but the books must stay in the shop.

Fire is its Kyptonite. The shop "dies" as it burns down.