Watamote 161, another fluffy chapter, but also laying the groundwork for potential future drama
- Ucchi practically just confessed to liking Tomoko, except that nobody else can translate Ucchi-ese. But she was at least able to clearly apologize to Tomoko.
- Really Ucchi is kind of interesting as a romantic character - I don't think most people are really rooting for her to get with Tomoko, considering she barely even actually knows Tomoko, it's a very high school kind of crush. The big hope is rather that Ucchi reaches self acceptance and can be open with herself and to her friends about her liking girls. And then she can head off to college and get her gay on.
- Asuka, as unreadable as ever, is either playing the peacemaker between the two of them, or inserting herself into the conversation as class mom, "Hey, stop messing with Tomoko."
- Mako is, unsurprisingly at this point, sitting with Yoshida, which means she's finally introduced to Mako's delinquent friends. At this point I'm all aboard the Mako/Yoshida ship... Even if not, I really like that it's not just Tomoko's friendships that keep developing, people are developing friendships with others as well, like Nemo and Yuri have really fun interactions. The interesting part with Mako and Yoshida is that the manga's been dropping a lot of hints that they've become pretty close and talk all the time, and also that nobody else in their group of friends knows about it.
- While everyone else is talking Yuri, classic introvert that she is, just goes ahead and silently eats all of the food.
- Asuka's curiousity/rivalry about Yuu has been building for a few chapters now, it'll be interesting to see where this develops. Yuu's the girl Tomoko used to sexually harass in middle school, her old clueless straight-girl crush, and still her closest friend. Asuka has been picking up on that and really seems on fire to find out what exactly their relationship is, and it's awfully hard to not read it as a feeling of romantic rivalry on Asuka's part.
Kana on
A trap is for fish: when you've got the fish, you can forget the trap. A snare is for rabbits: when you've got the rabbit, you can forget the snare. Words are for meaning: when you've got the meaning, you can forget the words.
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
0
Options
MaddocI'm Bobbin Threadbare, are you my mother?Registered Userregular
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
Yeah that's probably how this tech will end up being used, isn't it, bleh.
I was thinking it could be used to enhance existing production processes with additional inbetweens on top of the normal human made ones but that's totally not what it would be used for, they'd just fire the human inbetweeners.
and given that these are raw results from two years ago, the animation tech seems like its getting closer too
Teaching a computer how to smear seems like you could get 80% of the way which would be better than most 3s animation. Still wont have the style and fluidity of 1s but its definitely worth doing.
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Better art than us
I have a podcast now. It's about video games and anime!Find it here.
I also can't really stand Food Wars' cringey orgasm scenes. if you want some non-embarrassing food anime where the term food porn is more non-literal, I've seen a few good ones on crunchyroll. Today's Menu For The Emiya Family and Gourmet Girl Graffiti come to mind.
Wakakozake is also great. It's also very, very short : (
I watched the first episode or two, and it didn't click with me, seemed way too meandering. That is a problem I have with a lot of japanese media though
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Not necessarily
You load an algorithm up with enough rules for story structure, art design, and animation principles and there's no reason that it has to be intelligent to spit something out that's recognizable as a complete product
That sounds fuckin hard and even with all the progress people are talking about in here it's probably a good ways away, but it seems hypothetically possible
The main problem i would think is that it would struggle to produce things that don't come across as "formulaic," though I guess if you just trained it up on things that are considered "experimental" it might be able to? Iunno I think about how incoherent autocomplete messages are and think that anything like that must surely be far enough away that there's not much sense worrying about it yet
So... Quintessential Quintuplets. I've made it through 4 episodes, does it get any better or at least not just be "a harem, but the girls are all related!"?
So... Quintessential Quintuplets. I've made it through 4 episodes, does it get any better or at least not just be "a harem, but the girls are all related!"?
Of course not
I have a podcast now. It's about video games and anime!Find it here.
So... Quintessential Quintuplets. I've made it through 4 episodes, does it get any better or at least not just be "a harem, but the girls are all related!"?
Oh it’s harem garbage but the hook is admittedly good since they can all easily disguise as each other, the shitty dude met one or more of them years ago when they were all still dressing the same but doesn’t realize it, and we know he ends up married to one of them but no clue which.
That is not an endorsement of the show because it’s still empty calories at best, trashy at worst, and yet another harem where a bland shitty dude somehow gets all these wildly different girls to fall for him, but the hook is solid.
I finally figured out what it is about that first episode of MOMRPG that I don't like. They're pulling the usual bad-writing stunt of making it glaringly obvious to the audience what's actually going on (the whole isekai situation is deliberately engineered to help parents bond with emotionally-distant children), but despite Mamako being an absolute blabbermouth the son doesn't catch on. The plot might move past that phase, but the benefit of the doubt doesn't go far when youre dealing with fad genres.
My favorite musical instrument is the air-raid siren.
+1
Options
BroloBroseidonLord of the BroceanRegistered Userregular
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Not necessarily
You load an algorithm up with enough rules for story structure, art design, and animation principles and there's no reason that it has to be intelligent to spit something out that's recognizable as a complete product
That sounds fuckin hard and even with all the progress people are talking about in here it's probably a good ways away, but it seems hypothetically possible
The main problem i would think is that it would struggle to produce things that don't come across as "formulaic," though I guess if you just trained it up on things that are considered "experimental" it might be able to? Iunno I think about how incoherent autocomplete messages are and think that anything like that must surely be far enough away that there's not much sense worrying about it yet
style transfer is already something that neural networks can do, but they're not particularly adapted to 2D animation principles yet
doing some thing for in-betweening based on very similar keyframes would likely be easier though, since the goal of animation is to make a lot of frames of animation with very minute changes
you'd still need someone supervising and cleaning up the AI output, but it would still eliminate a lot of labor
is that a good thing? I'm not sure - animation is super time intensive when done well, and you get reports from studios like madhouse saying that they're being overworked in sweatshop like conditions to meet deadlines. but our track record with automating this kind of stuff usually results in the labor force being replaced by machines, and then increasing production to the point that all the remaining workers are overloaded and overworked anyway
I finally figured out what it is about that first episode of MOMRPG that I don't like. They're pulling the usual bad-writing stunt of making it glaringly obvious to the audience what's actually going on (the whole isekai situation is deliberately engineered to help parents bond with emotionally-distant children), but despite Mamako being an absolute blabbermouth the son doesn't catch on. The plot might move past that phase, but the benefit of the doubt doesn't go far when youre dealing with fad genres.
"I finally figured out what it is about that first episode of MOMRPG that I don't like."
"first episode of MOMRPG"
"MOMRPG"
Oh my god this only exists because of a typo doesn't it.
FF XIV - Qih'to Furishu (on Siren), Battle.Net - Ilpala#1975
Switch - SW-7373-3669-3011
Fuck Joe Manchin
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Not necessarily
You load an algorithm up with enough rules for story structure, art design, and animation principles and there's no reason that it has to be intelligent to spit something out that's recognizable as a complete product
That sounds fuckin hard and even with all the progress people are talking about in here it's probably a good ways away, but it seems hypothetically possible
The main problem i would think is that it would struggle to produce things that don't come across as "formulaic," though I guess if you just trained it up on things that are considered "experimental" it might be able to? Iunno I think about how incoherent autocomplete messages are and think that anything like that must surely be far enough away that there's not much sense worrying about it yet
style transfer is already something that neural networks can do, but they're not particularly adapted to 2D animation principles yet
doing some thing for in-betweening based on very similar keyframes would likely be easier though, since the goal of animation is to make a lot of frames of animation with very minute changes
you'd still need someone supervising and cleaning up the AI output, but it would still eliminate a lot of labor
is that a good thing? I'm not sure - animation is super time intensive when done well, and you get reports from studios like madhouse saying that they're being overworked in sweatshop like conditions to meet deadlines. but our track record with automating this kind of stuff usually results in the labor force being replaced by machines, and then increasing production to the point that all the remaining workers are overloaded and overworked.
yeah it seems a lot more reasonable to imagine a short-term where all the tweens can be algorithmically generated and just the keyframes are hand made
which, in addition to the problems you pointed out, i imagine that would still probably result in lower quality for a pretty long time, and generally be used for cheaper stuff
kinda like how flash is/was used for cheap kids' shows, and like flash i'm sure you'd have certain studios that are really talented at turning out solid quality even with that production method
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Not necessarily
You load an algorithm up with enough rules for story structure, art design, and animation principles and there's no reason that it has to be intelligent to spit something out that's recognizable as a complete product
That sounds fuckin hard and even with all the progress people are talking about in here it's probably a good ways away, but it seems hypothetically possible
The main problem i would think is that it would struggle to produce things that don't come across as "formulaic," though I guess if you just trained it up on things that are considered "experimental" it might be able to? Iunno I think about how incoherent autocomplete messages are and think that anything like that must surely be far enough away that there's not much sense worrying about it yet
I know a friend of mine who is a professional artist is convinced this will happen sooner or later.
I think I heard there was even an anime released recently about a future where all media is AI produced.
0
Options
ThegreatcowLord of All BaconsWashington State - It's Wet up here innit? Registered Userregular
I haven't watched MOMRPG/OkaasanOnline/Multi-TargetMom/MomIsekai, but I do love how many different nicknames it has
I finally tried to watch it this weekend and I'm struck by a line Gigguk did in reviewing a trend in recent romantic comedy animes that have been popping up as of late:
"I can't believe it's not hentai"
Good lord this show...just...there's thirst and and there's "are you living in the sahara and have run out of supplies 3 days ago?" I made it two episodes in and I felt skeevy by proxy. It also doesn't help that the good-intentions of the mother trying to bond with her son are twisted by the son being a complete wanker and them swan-diving hard into the implied incest route just turns the whole thing for me.
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Not necessarily
You load an algorithm up with enough rules for story structure, art design, and animation principles and there's no reason that it has to be intelligent to spit something out that's recognizable as a complete product
That sounds fuckin hard and even with all the progress people are talking about in here it's probably a good ways away, but it seems hypothetically possible
The main problem i would think is that it would struggle to produce things that don't come across as "formulaic," though I guess if you just trained it up on things that are considered "experimental" it might be able to? Iunno I think about how incoherent autocomplete messages are and think that anything like that must surely be far enough away that there's not much sense worrying about it yet
I know a friend of mine who is a professional artist is convinced this will happen sooner or later.
I think I heard there was even an anime released recently about a future where all media is AI produced.
That's Carole and Tuesday
A trap is for fish: when you've got the fish, you can forget the trap. A snare is for rabbits: when you've got the rabbit, you can forget the snare. Words are for meaning: when you've got the meaning, you can forget the words.
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Not necessarily
You load an algorithm up with enough rules for story structure, art design, and animation principles and there's no reason that it has to be intelligent to spit something out that's recognizable as a complete product
That sounds fuckin hard and even with all the progress people are talking about in here it's probably a good ways away, but it seems hypothetically possible
The main problem i would think is that it would struggle to produce things that don't come across as "formulaic," though I guess if you just trained it up on things that are considered "experimental" it might be able to? Iunno I think about how incoherent autocomplete messages are and think that anything like that must surely be far enough away that there's not much sense worrying about it yet
I know a friend of mine who is a professional artist is convinced this will happen sooner or later.
I think I heard there was even an anime released recently about a future where all media is AI produced.
i don't think it'd ever completely replace human made stuff
even if AI-driven animation and filmmaking could completely satisfy the demand for entertainment, it can't satisfy the desire to make things. It might simplify that process and make creation easier, like any other tool, but on some level people would want to express themselves. Kinda like how people still paint pictures of the real world even though cameras have been around for 200 years. there's no need for realistic paintings anymore, they could all go away tomorrow and most people probably wouldn't really mind that much if at all, but they're fulfilling to make so people keep making them anyway.
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Not necessarily
You load an algorithm up with enough rules for story structure, art design, and animation principles and there's no reason that it has to be intelligent to spit something out that's recognizable as a complete product
That sounds fuckin hard and even with all the progress people are talking about in here it's probably a good ways away, but it seems hypothetically possible
The main problem i would think is that it would struggle to produce things that don't come across as "formulaic," though I guess if you just trained it up on things that are considered "experimental" it might be able to? Iunno I think about how incoherent autocomplete messages are and think that anything like that must surely be far enough away that there's not much sense worrying about it yet
I know a friend of mine who is a professional artist is convinced this will happen sooner or later.
I think I heard there was even an anime released recently about a future where all media is AI produced.
i don't think it'd ever completely replace human made stuff
even if AI-driven animation and filmmaking could completely satisfy the demand for entertainment, it can't satisfy the desire to make things. It might simplify that process and make creation easier, like any other tool, but on some level people would want to express themselves. Kinda like how people still paint pictures of the real world even though cameras have been around for 200 years. there's no need for realistic paintings anymore, they could all go away tomorrow and most people probably wouldn't really mind that much if at all, but they're fulfilling to make so people keep making them anyway.
No, but it does shift the balance in good and bad ways. It makes it easier for small studios and creators to tell the stories they want to tell, but it also further increases competition for those few "big ticket" shows/movies that come out each year. So just a progression of what we're already seeing.
Using an algorithm to make 24 FPS into 60 FPS is going to look bad no matter what.
If you want something at 60 FPS make it natively at 60 FPS.
yeah they picked a pretty brute-force method of doing this that doesn't work well with hand animation
but at some point i wonder if ai will get good enough to do this
like with enough source footage you could probably train a deep learning intelligence to recognize a goku face/body and re-draw it from any angle
and given two clean key frames of animation you could probably do a decent job of guessing what an intermediate frame would look like instead of just blurring them together
as well as having bits of the ai to know when/where *not* to blend frames together by analyzing artifacts
you'd need to tune it a lot to pick up on artistic intent but it seems within the realm of reason
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
You're probably working with proper sapience at that point though, so at least that wouldn't be like, the death of art or somesuch.
Not necessarily
You load an algorithm up with enough rules for story structure, art design, and animation principles and there's no reason that it has to be intelligent to spit something out that's recognizable as a complete product
That sounds fuckin hard and even with all the progress people are talking about in here it's probably a good ways away, but it seems hypothetically possible
The main problem i would think is that it would struggle to produce things that don't come across as "formulaic," though I guess if you just trained it up on things that are considered "experimental" it might be able to? Iunno I think about how incoherent autocomplete messages are and think that anything like that must surely be far enough away that there's not much sense worrying about it yet
I know a friend of mine who is a professional artist is convinced this will happen sooner or later.
I think I heard there was even an anime released recently about a future where all media is AI produced.
i don't think it'd ever completely replace human made stuff
even if AI-driven animation and filmmaking could completely satisfy the demand for entertainment, it can't satisfy the desire to make things. It might simplify that process and make creation easier, like any other tool, but on some level people would want to express themselves. Kinda like how people still paint pictures of the real world even though cameras have been around for 200 years. there's no need for realistic paintings anymore, they could all go away tomorrow and most people probably wouldn't really mind that much if at all, but they're fulfilling to make so people keep making them anyway.
No, but it does shift the balance in good and bad ways. It makes it easier for small studios and creators to tell the stories they want to tell, but it also further increases competition for those few "big ticket" shows/movies that come out each year. So just a progression of what we're already seeing.
for sure.
like, depending on how sophisticated we're talking here, if for example Shonen Jump can open up a computer program and say, "give me a [ACTION MANGA] about a [HIGH SCHOOL BOY AGED 14-17] with [A SUPERPOWER] who [FIGHTS] [PTERODACTYLS] in order to [IMPRESS HIS] [LOVE INTEREST], drawn with [THICK] [EXPRESSIVE] lines and [DEEP SHADOWS]," and then just leave the computer to render panels overnight, then that'd completely change the economics of being a comic artist. It'd become something that only absolute superstars, who probably had established careers before the rise of that technology, could actually make a living doing.
but then that's true for a whole bunch of industries. Like, as soon as all the bugs with self-driving cars are worked out, then every single taxi driver, trucker, delivery driver is out of a job as soon as their employer can afford to buy a replacement for them. Capitalism is a shit system to begin with, but it gets less and less sensible the less and less scarce labor becomes. There's a tipping point at which either development of that kind of technology has to be completely abandoned or capitalism has to be replaced with something else (hopefully, but not necessarily, with something better).
I mean, we wouldn't have gotten One Punch Man or Mob Psycho 100 without the internet, so I'm prepared to sift through the ocean of bad to get to the new, actually good stuff. There's no danger of a content shortage anytime soon.
Posts
- Really Ucchi is kind of interesting as a romantic character - I don't think most people are really rooting for her to get with Tomoko, considering she barely even actually knows Tomoko, it's a very high school kind of crush. The big hope is rather that Ucchi reaches self acceptance and can be open with herself and to her friends about her liking girls. And then she can head off to college and get her gay on.
- Asuka, as unreadable as ever, is either playing the peacemaker between the two of them, or inserting herself into the conversation as class mom, "Hey, stop messing with Tomoko."
- Mako is, unsurprisingly at this point, sitting with Yoshida, which means she's finally introduced to Mako's delinquent friends. At this point I'm all aboard the Mako/Yoshida ship... Even if not, I really like that it's not just Tomoko's friendships that keep developing, people are developing friendships with others as well, like Nemo and Yuri have really fun interactions. The interesting part with Mako and Yoshida is that the manga's been dropping a lot of hints that they've become pretty close and talk all the time, and also that nobody else in their group of friends knows about it.
- While everyone else is talking Yuri, classic introvert that she is, just goes ahead and silently eats all of the food.
- Asuka's curiousity/rivalry about Yuu has been building for a few chapters now, it'll be interesting to see where this develops. Yuu's the girl Tomoko used to sexually harass in middle school, her old clueless straight-girl crush, and still her closest friend. Asuka has been picking up on that and really seems on fire to find out what exactly their relationship is, and it's awfully hard to not read it as a feeling of romantic rivalry on Asuka's part.
generously apply the techniques of squash and stretch
yeah its seeing how far (and disturbing) that stuff has come that makes me think it's a lot closer than i originally thought:
like this result isn't bad for a small team that had a weekend to create a fake
and given that these are raw results from two years ago, the animation tech seems like its getting closer too
well, you are now
Rock Band DLC | GW:OttW - arrcd | WLD - Thortar
Out of nowhere.
Total whiplash.
Not gonna lie, the fact that it's very possible AI could replace humans as the creators of media freaks me out.
I'm just imagining a future where everything from cartoons to "live action" movies are all completely produced by AI and all the actors aren't real.
I was thinking it could be used to enhance existing production processes with additional inbetweens on top of the normal human made ones but that's totally not what it would be used for, they'd just fire the human inbetweeners.
Teaching a computer how to smear seems like you could get 80% of the way which would be better than most 3s animation. Still wont have the style and fluidity of 1s but its definitely worth doing.
Better art than us
Wakakozake is also great. It's also very, very short : (
Listen to that music though!
https://www.crunchyroll.com/wakakozake-drama
Steam
Not necessarily
You load an algorithm up with enough rules for story structure, art design, and animation principles and there's no reason that it has to be intelligent to spit something out that's recognizable as a complete product
That sounds fuckin hard and even with all the progress people are talking about in here it's probably a good ways away, but it seems hypothetically possible
The main problem i would think is that it would struggle to produce things that don't come across as "formulaic," though I guess if you just trained it up on things that are considered "experimental" it might be able to? Iunno I think about how incoherent autocomplete messages are and think that anything like that must surely be far enough away that there's not much sense worrying about it yet
http://www.audioentropy.com/
Of course not
That is not an endorsement of the show because it’s still empty calories at best, trashy at worst, and yet another harem where a bland shitty dude somehow gets all these wildly different girls to fall for him, but the hook is solid.
style transfer is already something that neural networks can do, but they're not particularly adapted to 2D animation principles yet
doing some thing for in-betweening based on very similar keyframes would likely be easier though, since the goal of animation is to make a lot of frames of animation with very minute changes
you'd still need someone supervising and cleaning up the AI output, but it would still eliminate a lot of labor
is that a good thing? I'm not sure - animation is super time intensive when done well, and you get reports from studios like madhouse saying that they're being overworked in sweatshop like conditions to meet deadlines. but our track record with automating this kind of stuff usually results in the labor force being replaced by machines, and then increasing production to the point that all the remaining workers are overloaded and overworked anyway
"I finally figured out what it is about that first episode of MOMRPG that I don't like."
"first episode of MOMRPG"
"MOMRPG"
Oh my god this only exists because of a typo doesn't it.
Switch - SW-7373-3669-3011
Fuck Joe Manchin
Rock Band DLC | GW:OttW - arrcd | WLD - Thortar
yeah it seems a lot more reasonable to imagine a short-term where all the tweens can be algorithmically generated and just the keyframes are hand made
which, in addition to the problems you pointed out, i imagine that would still probably result in lower quality for a pretty long time, and generally be used for cheaper stuff
kinda like how flash is/was used for cheap kids' shows, and like flash i'm sure you'd have certain studios that are really talented at turning out solid quality even with that production method
http://www.audioentropy.com/
I know a friend of mine who is a professional artist is convinced this will happen sooner or later.
I think I heard there was even an anime released recently about a future where all media is AI produced.
I finally tried to watch it this weekend and I'm struck by a line Gigguk did in reviewing a trend in recent romantic comedy animes that have been popping up as of late:
"I can't believe it's not hentai"
Good lord this show...just...there's thirst and and there's "are you living in the sahara and have run out of supplies 3 days ago?" I made it two episodes in and I felt skeevy by proxy. It also doesn't help that the good-intentions of the mother trying to bond with her son are twisted by the son being a complete wanker and them swan-diving hard into the implied incest route just turns the whole thing for me.
Wud yoo laek to lern aboot meatz? Look here!
It was still good! I think I like S1 more though.
That's Carole and Tuesday
i don't think it'd ever completely replace human made stuff
even if AI-driven animation and filmmaking could completely satisfy the demand for entertainment, it can't satisfy the desire to make things. It might simplify that process and make creation easier, like any other tool, but on some level people would want to express themselves. Kinda like how people still paint pictures of the real world even though cameras have been around for 200 years. there's no need for realistic paintings anymore, they could all go away tomorrow and most people probably wouldn't really mind that much if at all, but they're fulfilling to make so people keep making them anyway.
http://www.audioentropy.com/
No, but it does shift the balance in good and bad ways. It makes it easier for small studios and creators to tell the stories they want to tell, but it also further increases competition for those few "big ticket" shows/movies that come out each year. So just a progression of what we're already seeing.
I think S2 has higher emotional highs than S2 for me, but S1 has a better overall arc, I guess?
But man, both seasons are still so good.
for sure.
like, depending on how sophisticated we're talking here, if for example Shonen Jump can open up a computer program and say, "give me a [ACTION MANGA] about a [HIGH SCHOOL BOY AGED 14-17] with [A SUPERPOWER] who [FIGHTS] [PTERODACTYLS] in order to [IMPRESS HIS] [LOVE INTEREST], drawn with [THICK] [EXPRESSIVE] lines and [DEEP SHADOWS]," and then just leave the computer to render panels overnight, then that'd completely change the economics of being a comic artist. It'd become something that only absolute superstars, who probably had established careers before the rise of that technology, could actually make a living doing.
but then that's true for a whole bunch of industries. Like, as soon as all the bugs with self-driving cars are worked out, then every single taxi driver, trucker, delivery driver is out of a job as soon as their employer can afford to buy a replacement for them. Capitalism is a shit system to begin with, but it gets less and less sensible the less and less scarce labor becomes. There's a tipping point at which either development of that kind of technology has to be completely abandoned or capitalism has to be replaced with something else (hopefully, but not necessarily, with something better).
http://www.audioentropy.com/
Gah, and now I'm all caught up and have to wait for more story, boooooooo.
Question about S2