From 45b6d85bdd656d7f5efb26d43a6a716546d46cfc Mon Sep 17 00:00:00 2001 From: Volpeon Date: Wed, 11 Jan 2023 08:39:55 +0100 Subject: Added a 404 page --- Makefile | 1 + content/404/home.md | 5 ++ content/404/index.md | 7 +++ content/notebook/archive/ai-art.md | 66 ++++++++++++++++++++++ content/notebook/archive/i-want-to-be-a-fox.md | 59 +++++++++++++++++++ .../archive/in-depth-image-descriptions.md | 31 ++++++++++ content/notebook/archive/index.md | 9 +++ content/notebook/fediverse/ai-art.md | 66 ---------------------- content/notebook/fediverse/i-want-to-be-a-fox.md | 59 ------------------- .../fediverse/in-depth-image-descriptions.md | 31 ---------- content/notebook/fediverse/index.md | 9 --- 11 files changed, 178 insertions(+), 165 deletions(-) create mode 100644 content/404/home.md create mode 100644 content/404/index.md create mode 100644 content/notebook/archive/ai-art.md create mode 100644 content/notebook/archive/i-want-to-be-a-fox.md create mode 100644 content/notebook/archive/in-depth-image-descriptions.md create mode 100644 content/notebook/archive/index.md delete mode 100644 content/notebook/fediverse/ai-art.md delete mode 100644 content/notebook/fediverse/i-want-to-be-a-fox.md delete mode 100644 content/notebook/fediverse/in-depth-image-descriptions.md delete mode 100644 content/notebook/fediverse/index.md diff --git a/Makefile b/Makefile index 1a7fae7..5eb2df1 100644 --- a/Makefile +++ b/Makefile @@ -191,6 +191,7 @@ out/assets/icons.svg: | out/assets graph \ git-branch \ smile \ + home \ link \ x \ chevron-left \ diff --git a/content/404/home.md b/content/404/home.md new file mode 100644 index 0000000..0758ed1 --- /dev/null +++ b/content/404/home.md @@ -0,0 +1,5 @@ +--- +title: Home page +icon: home +url: / +--- diff --git a/content/404/index.md b/content/404/index.md new file mode 100644 index 0000000..ff1745f --- /dev/null +++ b/content/404/index.md @@ -0,0 +1,7 @@ +--- +title: 404 +list_layout: grid-2 +unlisted: true +--- + +The page you wanted to visit does not exist. diff --git a/content/notebook/archive/ai-art.md b/content/notebook/archive/ai-art.md new file mode 100644 index 0000000..f8a9fd5 --- /dev/null +++ b/content/notebook/archive/ai-art.md @@ -0,0 +1,66 @@ +--- +schema_type: DigitalDocument +title: On AI Art +date: 2022-09-30 +last_update: 2022-11-13 + +references: + - label: "2022-09-30 -- Reply to my original post" + url: https://merveilles.town/@jbauer/109088036845654325 + - label: "2022-09-30 -- My Response" + url: https://mk.vulpes.one/notes/95s8h9ajtp + - label: "2022-11-13" + url: https://mk.vulpes.one/notes/97i33e337z + - label: "2022-11-13 -- Update 1" + url: https://mk.vulpes.one/notes/97ists0630 + - label: "2022-11-14" + url: https://mk.vulpes.one/notes/97k2fry860 +--- + +## 2022-09-30 -- Artists Weren't Happy When Photography Was Invented, Too + +::: alert +I posted the title on fedi. Someone replied and I elaborated on my views with the following post. +::: + +The primary reason AI art is widely getting banned is because a lot of people are wowed by the novelty of this technology and post their shitty results. But there are also people who spend a lot of time on tuning their prompts, running the results through img2img several times and touching things up manually in Photoshop. They're spending real effort on getting good results. + +AI is making good art a lot more accessible for everyone, including artists. I could run a drawing of mine through it to see what lighting or background works well, without having to spend hours on doing it myself and still not getting close to the quality of ""real"" artists. +People who have amazing ideas but lack the skills to draw them themselves finally have a tool that works for them. + +Also regarding sentience, this topic is so complex I don't even know where to begin, but I've made an interesting observation: The way this AI works isn't actually much different from us humans. + +Artists often use references and look up to other artists and adapt qualities from them in their own art. The way artist A draws scenery, the shading technique from B and C, and character designs from D -- of course not straight up copied and with clear boundaries, it all flows together as they see more art and feel inspired to adopt some qualities from it. This statement is true for me as well. + +Stable Diffusion was trained by looking at art as well. It doesn't have a database of every single picture. Instead it recognizes various concepts and puts them in some N-dimensional space -- it's memories. So one point in this space captures one concept it has seen in many different pictures. The prompt simply determines which of these concepts the AI will try to use (up to 75 with SD). + +Do you see the parallels? At least from my limited understanding, this doesn't seem much different from humans looking at pictures and recognizing shading, lighting, painting,... techniques and selectively using them in their own art. + +## 2022-11-13 -- The Concept of "Lost Commissions" + +An argument I hear often is that using an AI means I can get a piece in an artist's style when it should've been a commission. This doesn't really work for my situation because I don't copy an individual artist's style in my generated art -- it's a remix of things from both the original data and the data I added. I don't think commissioning ~10 artists at once is even possible, and if it was you'd better have lots of money. + +> Then commission an artist whose style is close enough + +Sure, I guess I could do that.\ +But this is where the second problem comes in: The way this usually goes with good artists (who are usually also popular) is that commissions are closed, the queues are full for the next 3 years, they never respond, and/or you have to keep track of them on Twitter for those 5 nanoseconds they're open before 10 billion followers snatch all slots.\ +I went through it all and I hated it. I won't even try commissioning an artist anymore if I notice any of these things because this is too stressful. + +And now there is AI that allows me to create good art of my random ideas. It's art that otherwise never would've existed because without the instantaneous feedback of AI I wouldn't even know if the idea worked or if I liked the implementation.\ +On the flip side, with ideas that proved to be good such as the fox-corvid-gryphon, I am now considering creating more art of it with my own abilities and maybe commission someone if there ever is a chance. So if anything, AI has *increased* my willingness to buy commissions. + +This whole situation is strikingly similar to the music/video/gaming industry, piracy and the concept of "lost sales". Food for thought. + +### Update 1 + +One more thing I got to experience when commissioning an artist: getting pushed back in the queue. 2 months turned into 2 years, and the sad thing is others have made even worse experiences (not necessarily with the same artist). + +There were probably things happening in the artist's life, or they possibly ran out of motivation which is something I can strongly relate to, or maybe they just suck at time management. I don't know. But it's undeniable that all these conditions are barriers. + +## 2022-11-14 -- AI is a Tool, not an Artist + +The mistake a lot of people make in this debate is that they're giving the AI the role of an artist rather than a tool. Stable Diffusion only works in conjunction with a human who operates it, and ultimately it is the human who decides which concepts to use and how via a prompt or more advanced techniques. It is still the human who brings in the "beyond" you were talking about. +It would be strange to apply standards of creativity to a pen or Photoshop, wouldn't it? + +Like any tool, AI can be used and abused. I can take Photoshop and plagiarize an existing work, and I can take an AI and plagiarize existing work. +In fact, people had the same concerns with the rise of digital art and later Photoshop because they made creating art so much easier. Yet here we are, digital art is common and Photoshop is an integral part of many artist's workflow. This is just another round of the same old debate. diff --git a/content/notebook/archive/i-want-to-be-a-fox.md b/content/notebook/archive/i-want-to-be-a-fox.md new file mode 100644 index 0000000..6fa75c7 --- /dev/null +++ b/content/notebook/archive/i-want-to-be-a-fox.md @@ -0,0 +1,59 @@ +--- +schema_type: DigitalDocument +title: I Want To Be a Fox +date: 2022-07-09 +last_update: 2022-07-30 + +references: + - label: Question + url: https://brotka.st/notice/ALFjUBS9tkCWKyprMm + - label: My Reply + url: https://mk.vulpes.one/notes/92gm6ipdtp + - label: Update 2022-07-30 + url: https://mk.vulpes.one/notes/93ajjhu562 +--- + +::: alert +@kaia@brotka.st asked: "maybe not something you want to answer at all or publicly, but I'd be curious whether you genuinely long to be a dragon or a fox, its intensity, the motivation and background" +::: + +I don't mind replying in public. +It's not like I'm trying to keep this a secret, I just rarely talk about it because I usually have nothing to say on this matter. :drgn_flat: + +So yeah, I do genuinely wish to be a fox. +I don't know the source of this desire, I've just always felt more comfortable with the idea of being an animal than a human. +Some of my earliest memories are about me imagining myself as a bird and focusing on the sensation of having such a different body.\ +It's been a constant throughout my life and I can hardly express how much I long for it to become reality. +It doesn't matter that I'm completely aware of all the drawbacks and consequences, if I was given the chance I would take it in a heartbeat. + +Obviously, that will never happen, though. +I accepted it because what other choice do I have, but I'd be lying if I said that it wasn't eating away at me at times. +I found two things that help me handle my feelings: The first one is creating art. +The second one is my online identity which lets me live out myself genuinely, unlike my "real" name and my "real" face. +Here I'm not strictly limited to my body and I'm truly thankful that this possibility exists. + +As for the recent dragon stuff, I thought about what this character means to me. +I identify with it because it's an expression of my creativity and taste, but it hasn't taken over the role of foxes. + +## Update 2022-07-30 + +::: alert +I recent post that I feel is relevant here. +::: + +I mentioned before that I sometimes have sleepwalking...ish episodes where I partially wake up at night and I'm usually aware of who and where I am, but my mind is still in dream mode which leads to interesting thought processes. + +For instance, about half a year ago I had 3 or 4 where I thought wasps (closest approximation) were attacking me. +I'm sure I posted about them on fedi. +It fucking sucked. + +Last night I had one where I thought I was a spirit and I could take any shape I wanted, so naturally I tried to be a fox. +Then I fully woke up because my heart was racing and I had to calm down.\ +It's not the first time an episode went in that direction and I cherish every time it happens. +In these rare moments, being a fox is suddenly a dream that can become reality right then and there and I experience a peace of mind that's indescribable. +Just try to imagine your feelings if an impossible lifelong dream suddenly came true.\ +Nothing else comes even close to this experience, because when I'm awake I'm always aware of the reality in the back of my mind. +No amount of fantasizing or online interactions or well-intentioned "but Volpeon, you _are_ a fox"-es or anything else can change that. +Which doesn't mean it can't be helpful -- except for the last one in a serious context, I don't like that at all. + +Anyway, last night was great. :drgn_happy: diff --git a/content/notebook/archive/in-depth-image-descriptions.md b/content/notebook/archive/in-depth-image-descriptions.md new file mode 100644 index 0000000..3143a36 --- /dev/null +++ b/content/notebook/archive/in-depth-image-descriptions.md @@ -0,0 +1,31 @@ +--- +schema_type: DigitalDocument +title: Detailed Image Descriptions Aren't Always Helpful +date: 2022-05-05 + +references: + - label: Original Post + url: https://mk.vulpes.one/notes/8zwxp9lywc +--- + +::: alert +I noticed that quite a few people write very extensive image descriptions. +Mastodon in particular implies that this is the correct way since the character limit for them is [three times as high as the limit for the actual post content](https://github.com/mastodon/mastodon/pull/11819). +However, there are common situations where these image descriptions are in fact not helpful. +::: + +Imagine you're using an online shop. +What you'll typically see is a list of products where each list item has a small preview image, a product name and some other metadata.\ +Now imagine that the shop displays a long textual description for each product instead. +This is what the timeline appears like to screen reader users if long image descriptions are used. +And unlike people seeing the text, they don't have the luxury of skimming to grasp vital information quickly -- they can only wait for the screen reader to read all of it or until they decide to skip the rest. + +Detailed descriptions are only helpful when the user has decided the content is interesting to them. +Currently, most fedi frontends put them in the attachment's `alt` attribute, which is fine if the user is currently viewing a single post instead of the timeline. +But on the timeline, it's much more important to have quick summaries instead.\ +In terms of the example above, it's the same when viewing a single product vs. the product list. +You're only interested in the details if a product is interesting to you. + +Image description on fedi are a step up from having none at all, but they're still inadequate. +Ideally, you could provide both a quick summary and an in-depth description, and the UI presents both in a way that's the most helpful.\ +Since this possibility doesn't exist, please consider keeping image descriptions short and putting the long description in the post body. diff --git a/content/notebook/archive/index.md b/content/notebook/archive/index.md new file mode 100644 index 0000000..8892637 --- /dev/null +++ b/content/notebook/archive/index.md @@ -0,0 +1,9 @@ +--- +title: Archive +position: 2 +list_order: date_desc +list_read_indicators: true +feed: true +--- + +An archive of my posts from other platforms that I wanted to preserve. diff --git a/content/notebook/fediverse/ai-art.md b/content/notebook/fediverse/ai-art.md deleted file mode 100644 index f8a9fd5..0000000 --- a/content/notebook/fediverse/ai-art.md +++ /dev/null @@ -1,66 +0,0 @@ ---- -schema_type: DigitalDocument -title: On AI Art -date: 2022-09-30 -last_update: 2022-11-13 - -references: - - label: "2022-09-30 -- Reply to my original post" - url: https://merveilles.town/@jbauer/109088036845654325 - - label: "2022-09-30 -- My Response" - url: https://mk.vulpes.one/notes/95s8h9ajtp - - label: "2022-11-13" - url: https://mk.vulpes.one/notes/97i33e337z - - label: "2022-11-13 -- Update 1" - url: https://mk.vulpes.one/notes/97ists0630 - - label: "2022-11-14" - url: https://mk.vulpes.one/notes/97k2fry860 ---- - -## 2022-09-30 -- Artists Weren't Happy When Photography Was Invented, Too - -::: alert -I posted the title on fedi. Someone replied and I elaborated on my views with the following post. -::: - -The primary reason AI art is widely getting banned is because a lot of people are wowed by the novelty of this technology and post their shitty results. But there are also people who spend a lot of time on tuning their prompts, running the results through img2img several times and touching things up manually in Photoshop. They're spending real effort on getting good results. - -AI is making good art a lot more accessible for everyone, including artists. I could run a drawing of mine through it to see what lighting or background works well, without having to spend hours on doing it myself and still not getting close to the quality of ""real"" artists. -People who have amazing ideas but lack the skills to draw them themselves finally have a tool that works for them. - -Also regarding sentience, this topic is so complex I don't even know where to begin, but I've made an interesting observation: The way this AI works isn't actually much different from us humans. - -Artists often use references and look up to other artists and adapt qualities from them in their own art. The way artist A draws scenery, the shading technique from B and C, and character designs from D -- of course not straight up copied and with clear boundaries, it all flows together as they see more art and feel inspired to adopt some qualities from it. This statement is true for me as well. - -Stable Diffusion was trained by looking at art as well. It doesn't have a database of every single picture. Instead it recognizes various concepts and puts them in some N-dimensional space -- it's memories. So one point in this space captures one concept it has seen in many different pictures. The prompt simply determines which of these concepts the AI will try to use (up to 75 with SD). - -Do you see the parallels? At least from my limited understanding, this doesn't seem much different from humans looking at pictures and recognizing shading, lighting, painting,... techniques and selectively using them in their own art. - -## 2022-11-13 -- The Concept of "Lost Commissions" - -An argument I hear often is that using an AI means I can get a piece in an artist's style when it should've been a commission. This doesn't really work for my situation because I don't copy an individual artist's style in my generated art -- it's a remix of things from both the original data and the data I added. I don't think commissioning ~10 artists at once is even possible, and if it was you'd better have lots of money. - -> Then commission an artist whose style is close enough - -Sure, I guess I could do that.\ -But this is where the second problem comes in: The way this usually goes with good artists (who are usually also popular) is that commissions are closed, the queues are full for the next 3 years, they never respond, and/or you have to keep track of them on Twitter for those 5 nanoseconds they're open before 10 billion followers snatch all slots.\ -I went through it all and I hated it. I won't even try commissioning an artist anymore if I notice any of these things because this is too stressful. - -And now there is AI that allows me to create good art of my random ideas. It's art that otherwise never would've existed because without the instantaneous feedback of AI I wouldn't even know if the idea worked or if I liked the implementation.\ -On the flip side, with ideas that proved to be good such as the fox-corvid-gryphon, I am now considering creating more art of it with my own abilities and maybe commission someone if there ever is a chance. So if anything, AI has *increased* my willingness to buy commissions. - -This whole situation is strikingly similar to the music/video/gaming industry, piracy and the concept of "lost sales". Food for thought. - -### Update 1 - -One more thing I got to experience when commissioning an artist: getting pushed back in the queue. 2 months turned into 2 years, and the sad thing is others have made even worse experiences (not necessarily with the same artist). - -There were probably things happening in the artist's life, or they possibly ran out of motivation which is something I can strongly relate to, or maybe they just suck at time management. I don't know. But it's undeniable that all these conditions are barriers. - -## 2022-11-14 -- AI is a Tool, not an Artist - -The mistake a lot of people make in this debate is that they're giving the AI the role of an artist rather than a tool. Stable Diffusion only works in conjunction with a human who operates it, and ultimately it is the human who decides which concepts to use and how via a prompt or more advanced techniques. It is still the human who brings in the "beyond" you were talking about. -It would be strange to apply standards of creativity to a pen or Photoshop, wouldn't it? - -Like any tool, AI can be used and abused. I can take Photoshop and plagiarize an existing work, and I can take an AI and plagiarize existing work. -In fact, people had the same concerns with the rise of digital art and later Photoshop because they made creating art so much easier. Yet here we are, digital art is common and Photoshop is an integral part of many artist's workflow. This is just another round of the same old debate. diff --git a/content/notebook/fediverse/i-want-to-be-a-fox.md b/content/notebook/fediverse/i-want-to-be-a-fox.md deleted file mode 100644 index 6fa75c7..0000000 --- a/content/notebook/fediverse/i-want-to-be-a-fox.md +++ /dev/null @@ -1,59 +0,0 @@ ---- -schema_type: DigitalDocument -title: I Want To Be a Fox -date: 2022-07-09 -last_update: 2022-07-30 - -references: - - label: Question - url: https://brotka.st/notice/ALFjUBS9tkCWKyprMm - - label: My Reply - url: https://mk.vulpes.one/notes/92gm6ipdtp - - label: Update 2022-07-30 - url: https://mk.vulpes.one/notes/93ajjhu562 ---- - -::: alert -@kaia@brotka.st asked: "maybe not something you want to answer at all or publicly, but I'd be curious whether you genuinely long to be a dragon or a fox, its intensity, the motivation and background" -::: - -I don't mind replying in public. -It's not like I'm trying to keep this a secret, I just rarely talk about it because I usually have nothing to say on this matter. :drgn_flat: - -So yeah, I do genuinely wish to be a fox. -I don't know the source of this desire, I've just always felt more comfortable with the idea of being an animal than a human. -Some of my earliest memories are about me imagining myself as a bird and focusing on the sensation of having such a different body.\ -It's been a constant throughout my life and I can hardly express how much I long for it to become reality. -It doesn't matter that I'm completely aware of all the drawbacks and consequences, if I was given the chance I would take it in a heartbeat. - -Obviously, that will never happen, though. -I accepted it because what other choice do I have, but I'd be lying if I said that it wasn't eating away at me at times. -I found two things that help me handle my feelings: The first one is creating art. -The second one is my online identity which lets me live out myself genuinely, unlike my "real" name and my "real" face. -Here I'm not strictly limited to my body and I'm truly thankful that this possibility exists. - -As for the recent dragon stuff, I thought about what this character means to me. -I identify with it because it's an expression of my creativity and taste, but it hasn't taken over the role of foxes. - -## Update 2022-07-30 - -::: alert -I recent post that I feel is relevant here. -::: - -I mentioned before that I sometimes have sleepwalking...ish episodes where I partially wake up at night and I'm usually aware of who and where I am, but my mind is still in dream mode which leads to interesting thought processes. - -For instance, about half a year ago I had 3 or 4 where I thought wasps (closest approximation) were attacking me. -I'm sure I posted about them on fedi. -It fucking sucked. - -Last night I had one where I thought I was a spirit and I could take any shape I wanted, so naturally I tried to be a fox. -Then I fully woke up because my heart was racing and I had to calm down.\ -It's not the first time an episode went in that direction and I cherish every time it happens. -In these rare moments, being a fox is suddenly a dream that can become reality right then and there and I experience a peace of mind that's indescribable. -Just try to imagine your feelings if an impossible lifelong dream suddenly came true.\ -Nothing else comes even close to this experience, because when I'm awake I'm always aware of the reality in the back of my mind. -No amount of fantasizing or online interactions or well-intentioned "but Volpeon, you _are_ a fox"-es or anything else can change that. -Which doesn't mean it can't be helpful -- except for the last one in a serious context, I don't like that at all. - -Anyway, last night was great. :drgn_happy: diff --git a/content/notebook/fediverse/in-depth-image-descriptions.md b/content/notebook/fediverse/in-depth-image-descriptions.md deleted file mode 100644 index 3143a36..0000000 --- a/content/notebook/fediverse/in-depth-image-descriptions.md +++ /dev/null @@ -1,31 +0,0 @@ ---- -schema_type: DigitalDocument -title: Detailed Image Descriptions Aren't Always Helpful -date: 2022-05-05 - -references: - - label: Original Post - url: https://mk.vulpes.one/notes/8zwxp9lywc ---- - -::: alert -I noticed that quite a few people write very extensive image descriptions. -Mastodon in particular implies that this is the correct way since the character limit for them is [three times as high as the limit for the actual post content](https://github.com/mastodon/mastodon/pull/11819). -However, there are common situations where these image descriptions are in fact not helpful. -::: - -Imagine you're using an online shop. -What you'll typically see is a list of products where each list item has a small preview image, a product name and some other metadata.\ -Now imagine that the shop displays a long textual description for each product instead. -This is what the timeline appears like to screen reader users if long image descriptions are used. -And unlike people seeing the text, they don't have the luxury of skimming to grasp vital information quickly -- they can only wait for the screen reader to read all of it or until they decide to skip the rest. - -Detailed descriptions are only helpful when the user has decided the content is interesting to them. -Currently, most fedi frontends put them in the attachment's `alt` attribute, which is fine if the user is currently viewing a single post instead of the timeline. -But on the timeline, it's much more important to have quick summaries instead.\ -In terms of the example above, it's the same when viewing a single product vs. the product list. -You're only interested in the details if a product is interesting to you. - -Image description on fedi are a step up from having none at all, but they're still inadequate. -Ideally, you could provide both a quick summary and an in-depth description, and the UI presents both in a way that's the most helpful.\ -Since this possibility doesn't exist, please consider keeping image descriptions short and putting the long description in the post body. diff --git a/content/notebook/fediverse/index.md b/content/notebook/fediverse/index.md deleted file mode 100644 index 5e9f97c..0000000 --- a/content/notebook/fediverse/index.md +++ /dev/null @@ -1,9 +0,0 @@ ---- -title: Fediverse -position: 2 -list_order: date_desc -list_read_indicators: true -feed: true ---- - -An archive of my posts on the fediverse that I wanted to preserve. -- cgit v1.2.3-70-g09d2