Well some of them are slightly better, but I reached that conclusion way sooner than you so I cannot express myself on the more recent ones. Our tastes seem to coincide a lot, sadly I am not the best to recomend games but i seem to be good at pointing which ones to avoid, that may just be a matter of statistics however.
Thanks for the criticism. I wrote the last phrase just to receive posts like this. You took the time to write such a wealth of text, so I would also like to respond to your work.
Initially, I planned to use AI to write text for scenes, but I quickly moved away from this idea, because... the text turned out to be quite dry. Therefore, the text and thoughts are my own. And I understand that the text I wrote may be difficult to read because... This is my first significant experience in writing.
As for what the ship city looks like, I think it can look whatever you want.
I also admit that perhaps I could not cut out the characters everywhere normally, but I also wanted you to understand that creating images using SD is not as easy a job as it might seem at first glance. Working on an image takes a lot longer than I could create in a 3D platform. But thanks to this, the image turns out to be more 2d. Initially, I wanted to create images at the level of 2D art, but I lost track of this idea along the way. I'm currently working on simplifying the process in order to create quality images.
Regarding the similarities with LIL, I took the main concept from there, but I want to tell my own story.
- Jan 3, 2021
- 1,065
- 3,365
Oh, you wrote that to fish for criticism, and not just to solicit content ideas from your audience?
Not sure I buy it. Is this actually true, or is this just an attempt at damage control? Why should an audience believe that someone who has demonstrated so little care and effort with their visuals, has instead by some miracle put actual work into the writing?
Was it more or less work than you put into translating the game into 5 different languages I wonder?
Can it look like whatever you want? Sure. That's how fiction work. If you were to actually have drawn or modeled your own original cityscapes, demonstrating that there was the tiniest spark of your own creative vision fueling that decision? Sure, have at it. You fly your Akihabara looking spaceship to wherever you can imagine with your head held high.
But, that is not what you did. Asking Stable Diffusion to generate urban landscapes is simply the path of least resistance.
For the record, you cannot even be bothered to take the time to perform a proper selection when you cut the girls out of their backgrounds, leaving ugly jagged edges around them. Even when farming out the majority of the work to the algorithm, you cannot put in the modicum of effort for a simple editing task; opting instead for an effortless auto-selection tool that chops out part of a girl's shoulder in the process. You didn't care enough to prevent such sloppy work in the first place, or even catch it later and fix it before release. These things are NOT hard to spot if you put forth just the smallest bit of effort.
Do you think that explaining how you are 'working on simplifying the process' is somehow impressive? That's just an admission that you're looking for ways to make things easier, to take even less effort than you've already shown. You've again demonstrated nothing but a desire to follow the path of least resistance.
You took the concept.
You took the MC having a voice in their head not their own.
You took the daily grind, event format.
You took the amnesiac isekai protagonist.
You took the pretentiously weird, head-stuck-up-ass writing.
You took the criminal abuse of the ellipses.
So, what did you actually bring to the table? Art assets made by an algorithm you didn't create, and subsequently mangled during editing? A script supposedly written by someone who openly admits to already trying to offload the work onto an algorithm, and has demonstrably done so with the visuals already?
Honestly, the most original thing about your project is the Patreon link, and that is perhaps the most telling thing of all...
- Apr 29, 2018
- 9,766
- 2,162
Don't wanna spread hate ofc but AI art still needs some work. I'm just a normal dude and not pixel peeping like some people do. But when I see AI art, I can almost with certainty tell it's not impressive. It looks complete at first glance but somehow lacks the "spirit" and when any person does art, even when its horrible, you can still be charmed by its uniqueness. I think, AI art will definitely be a part of our life and I'm all for it but I don't think it can replace the character's art ever. It can, however, design the background and stuff that we don't give much attention to. Within 5/10 years, we can replace the boring part of art to AI, and just do the ones that require the most work. Yet I always find myself asking if this is better?
Yes, I agree that most of the work done by AI may not be impressive. But there are also outstanding works among them, in which there is more than enough soul. Right now, the main problem with generational AI is that you can't get what you need for a long time, even despite a lot of tips. And you have to spend a lot of attempts on generation. There are also many other problems such as not understanding the perspective in the image, etc. But I think the situation should improve in the coming year. Generative models can already be used to generate a "base image" for further processing.
- Jan 3, 2021
- 1,065
- 3,365
'AI art has soul'
...but also...
'You need to roll the dice a lot with AI art to get something nice'
Listed problems with AI don't include the wholesale theft from actual artists the world over, or the unsustainable amount of electricity and fresh water that is required to run the data centers powering them, or the catastrophic levels of CO2 their use is pumping into the atmosphere. All for a parlor trick created by Silicon Valley Tech-Bros and fueled by Venture Capitalist Ghouls, the same combo that already brought us scams like Cryptocurrency and NFT's.
Well, dude, these are too far-fetched reasons not to like AI technology. I also don't like cryptocurrencies and everything related to them, because carbon emissions, etc. But cryptocurrencies help people in those countries where the government is unable to ensure a stable exchange rate of the currency.
After all, animals also produce harmful emissions into the atmosphere, but we don't ban them.
As for the fact that neural networks steal intellectual property, I can't agree at all. Just as a person copies everything that is given to him during education, the neural network does the same.
- Jan 3, 2021
- 1,065
- 3,365
They're perfectly valid reasons. You just need to care about more than your immediate self.
That's propaganda bullshit spread by Cryptobros to hype and green-wash their 'speculative investment' (read: gambling & fraud), which only makes them money when enough other people agree with them. It is all vacuous hype.
Honestly, we probably should. Factory farming the meat consumption habits of the developed world as they are is environmentally destructive, and will only get worse as demand increases.
Complete bullshit. Neural networks don't comprehend. They repeat patterns. They are not general purpose intelligences. They are not 'studying' art the way a person does. Also, that's not how people teach or learn art. My art teachers never gave me a copy of someone else's work and instructed me to just copy it as a means to practice. That you'd even equate the two is both insulting, and belies the fact that you have zero experience in the arts. You have no fucking clue, which seems par for the course. AI simps either are ignorant of art and artists or actively resent and hate them.
At the end of the day, they are a commercial product. They were created by wholesale scrapping of content from the internet, without compensation or attribution. They are built upon theft. The Techbros didn't stop and ask for permission, and the VC Ghouls sure as shit didn't want to pay for access, as either would add time and cost to their hype-train parlor trick.
If Stable Diffusion, MidJourney and such actually did 'learn' like a human, then their current usage would amount to slavery. So either the algorithm learns just like a human and there is no theft, but you've made use of digital slave labor with a trapped and shackled General Purpose Intelligence. Or you are using a commercial product, a black box algorithm which was absolutely created by using millions of other original works without permission or compensation. PICK ONE.
This shit offers zero utility over previous algorithms that already handled stuff like search engine queries, outside of filling the internet with exponentially more spam and enabling zero-effort grifters like yourself. Twitter is filled with ChatGPT powered blue checkmarks replying to each other about their 'pussy in bio'. Goggle image search is awash in generative AI images of nonsense. Amazon and Etsy are filled with listings and images offloaded onto AI. All this shit has done is slap rocket boosters to the side of the internet's already crumbling usability; all so some VC Ghouls can make a profit at the expense of actual creative people and the environment.
Last edited:
- Jan 12, 2022
- 4
- 12
Hey, I wanted to give my feedback.
Art style
For starters, the use of AI CG might be quite controversial, especially seeing the exchange above. I think that regrettably we'll probably see more of it in the future and I might as well try and get used to it early on. Nevertheless, I'll still judge it by the standards of hand-drawn art, as right now it's still superior to the AI one and it's, I'd say, a standard. So, the art is really inconsistent. If it wasn't for the hair colour, I wouldn't be able to say those aren't different characters. Especially Vis, her face and hairstyle changes dramatically in those few scenes she has. There are obviously some suspicious artifacts all over the arts that make it really obvious it's an AI art. Overall the graphical quality isn't very impressive and even disappointing at some points. Though, there are some nice CGs, but in most cases it looks good only when you squint your eyes.
Language
I saw some negative comments about the language that it looks like it is AI generated. As much as it is true that it doesn't sound very natural, I believe it is written by the author himself. I'm a Slavic language speaker and I saw some interference from those languages, especially gendering nouns was matching my language, e.g. "weather" was later referred to as "she", where in for example German it's neuter, not feminine. There were some cases that the text was hard to understand due to odd choice of words, but overall not so bad. I also think that in, what I presume are, events from the version 0.2.0 it was much better, so I'm counting on you that you'll improve with time, your language proficiency as well as overall writing skills.
Gameplay
Well, it is a standard gameplay, that looks like many other games in ren'py using koikatsu, but because of that I have a very important note. Please add something to UI that lets you keep track of progress! I assume there are two events per character, but I've got every one of them to 10 affection and still haven't got any events for Ashley beside the story events in her dormitory and in office. I suspect that those were those two events for her, but I have no way of knowing. I could also add that the gameplay is sometimes tedious, as you cannot get any affection with characters, which you can see in the weekends, and there are 7 of them out of 10!
Story
To be honest I was very skeptical at the beginning, as it is AI art, but I was positively surprised! This whole "AI controlling the spaceship and the only person who can control it is an amnesiac" thing really got me interested. It matches really well with the art style actually. However, the inspiration taken from LiL is a little bit too excessive. I mean I enjoyed playing LiL and some other games taking inspiration from it, but important for me is, that it is not copying and, as for now, there are too many things resembling LiL. Still, the AI motive is interesting and surely a step in the right direction. Having said that, I have one thing to point out in the story.
When it comes to characters, I enjoyed them. There are some satisfactorly written archetypes and some with quite good twists to their characters. If it is the first time, you have written any characters, then good job! Still regret we didn't get more Ashley.
Overall it's not a game I have much hopes on and I wouldn't be surprised, if the author just left the project after one or two patches more. As it is still a new and short game, I think I will play the next updates without expecting much. But I hope that you will prove me wrong, so good luck!
Art style
For starters, the use of AI CG might be quite controversial, especially seeing the exchange above. I think that regrettably we'll probably see more of it in the future and I might as well try and get used to it early on. Nevertheless, I'll still judge it by the standards of hand-drawn art, as right now it's still superior to the AI one and it's, I'd say, a standard. So, the art is really inconsistent. If it wasn't for the hair colour, I wouldn't be able to say those aren't different characters. Especially Vis, her face and hairstyle changes dramatically in those few scenes she has. There are obviously some suspicious artifacts all over the arts that make it really obvious it's an AI art. Overall the graphical quality isn't very impressive and even disappointing at some points. Though, there are some nice CGs, but in most cases it looks good only when you squint your eyes.
Language
I saw some negative comments about the language that it looks like it is AI generated. As much as it is true that it doesn't sound very natural, I believe it is written by the author himself. I'm a Slavic language speaker and I saw some interference from those languages, especially gendering nouns was matching my language, e.g. "weather" was later referred to as "she", where in for example German it's neuter, not feminine. There were some cases that the text was hard to understand due to odd choice of words, but overall not so bad. I also think that in, what I presume are, events from the version 0.2.0 it was much better, so I'm counting on you that you'll improve with time, your language proficiency as well as overall writing skills.
Gameplay
Well, it is a standard gameplay, that looks like many other games in ren'py using koikatsu, but because of that I have a very important note. Please add something to UI that lets you keep track of progress! I assume there are two events per character, but I've got every one of them to 10 affection and still haven't got any events for Ashley beside the story events in her dormitory and in office. I suspect that those were those two events for her, but I have no way of knowing. I could also add that the gameplay is sometimes tedious, as you cannot get any affection with characters, which you can see in the weekends, and there are 7 of them out of 10!
Story
To be honest I was very skeptical at the beginning, as it is AI art, but I was positively surprised! This whole "AI controlling the spaceship and the only person who can control it is an amnesiac" thing really got me interested. It matches really well with the art style actually. However, the inspiration taken from LiL is a little bit too excessive. I mean I enjoyed playing LiL and some other games taking inspiration from it, but important for me is, that it is not copying and, as for now, there are too many things resembling LiL. Still, the AI motive is interesting and surely a step in the right direction. Having said that, I have one thing to point out in the story.
You don't have permission to view the spoiler content.
Log in or register now.
Overall it's not a game I have much hopes on and I wouldn't be surprised, if the author just left the project after one or two patches more. As it is still a new and short game, I think I will play the next updates without expecting much. But I hope that you will prove me wrong, so good luck!
Thank you for such a detailed review. Feedback is important to me right now.
I was trying to make the maximum graphic identity of the characters. Perhaps Vis differs greatly from scene to scene, because I changed the way the frame was generated. In the future, differences in the appearance of the characters should be minimal.
You have correctly noticed that I write the text in Russian and only then translate it into English. I continue to improve my English, but sometimes I can make mistakes.
I agree that more ways to enhance relationships need to be added, I'll think about it in future updates. I also want to add hints for passing to the next version.
I also think I took too much from LIL) When I started writing the script framework, I was too influenced by Lil. But I think the thoughts that I put into this script and the thoughts that Selebus put into it are quite different.
You don't have permission to view the spoiler content.
Log in or register now.
Thanks again for your feedback !