Profile avatar
necrokuma.bsky.social
Artist starting on this platform, I hope you like my drawings NO AI! vgen.co/NecroKuma3
189 posts 544 followers 291 following
Regular Contributor
Active Commenter

like...writers...you can't rail against A/I books then turn around and use A/I covers artists, why would you denounce A/I art but casually use ch@tgpt same goes for A/I voices and music we are ALL at risk of getting exploited -- we shouldn't be sabotaging each other and making it easier for them

I'm so sorry I can't stop laughing. OpenAI, the company built on stealing literally the entire internet, is crying because DeepSeek may have trained on the outputs from ChatGPT. They're crying their eyes out. What a bunch of hypocritical little babies. Cry more, freaks. www.ft.com/content/a0df...

パブコメ対応 2025年2月23日 18時0分まで 「企業取引研究会報告書」に関する意見募集 public-comment.e-gov.go.jp/pcm/detail?C... 公正取引委員会・中小企業庁 今日の18時までです。 この報告書は、主に下請法の見直しを視野に「物価や賃金が構造的に上がっていく経済社会における適正な企業間取引の在り方を見据え、適正な取引が透明性の高い形で実現される環境整備について、検討する」ものです。 2 (4) 知的財産・ノウハウの取引適正化に関する論点 に、生成AIとは記載されていませんが、 生成AIの学習を許諾するかどうかは、知的財産の取引です。

I have almost 200 PSD of drawings that I never finished because I was too distrustful of my color skills, including this Tracer that I drew in 2016 (one of the few sketches I was able to save from my family erasing them of the old PC)

Google Scholar is filled with AI-generated junk science. “Two-thirds of the papers they studied were at least in part produced through undisclosed use of GPTs.”

Lawyers are funny. Saying ChatGPT does not plagiarize “Willy nilly” is a cute way of saying “but it *does* plagiarize.” I could open MS Word infinity times, and it will never accidentally open NYT articles. I could open Google Docs infinity times, and it will accidentally plagiarize NYT.

I did NOT activate this, I had it disabled, and yet, now it appears enabled again, that is why "opt out" is a fucking joke, they can decide when they take yours, and not you

Holy shit this filing update is wild! It states that: 1. Zuckerberg gave the ok to train off copyrighted works 2. Engineers wrote scripts to remove copyright info 3. Meta downloaded illegal torrents to train models of GenAi is a house of cards, I swear. techcrunch.com/2025/01/09/m...

There is even more to this case and not getting the necessary coverage: @kortizart.bsky.social @ednewtonrex.bsky.social @necrokuma.bsky.social @rahll.bsky.social @artificialbodies.net @stealcase.bsky.social hindustantimes.com/world-news/u... legacy.www.documentcloud.org/documents/25...

アナログでも色塗ったんだよね #NARUTO #イラスト

Work in Progress. Ice cream with toppings

When anyone told you that "AI learns like a human", remember that the "learning" process of an AI is basically this:

This sums it up perfectly. It’s not a conversation.

A terminology proposal:

Going from reading about AI to reading about neuroscience is interesting, especially because you see how the paths diverge further and further.

How bad it is to work from home and your family think that that means you don't work and that they can use your time for things they need...

I strongly hate Japanese politicians who do not understand what Japanese animation really needs to return to the golden age and the quality and talent of yesteryear. AI is not needed, what is really needed is budget, time, and above all teach techniques www.youtube.com/watch?v=pEa_...

This is a scam that’s prevalent on Bluesky so please be aware. The scammer will copy an artist’s profile & create a nearly identical handle. They block the artist & reach out to their followers asking for money. If you’re contacted by a scammer let the artist know then report the scammer please.

I would like to make a comic explaining to artists in simple terms how convolutional networks, perceptrons and consequently any DNN-based REgenerative models such as diffusion models and LLMs work, something for artists to use as an argument for the classic "learn like a human" argument.

I'm weird...I can't stay drawing one style without getting bored, is this normal?

今日になってしまいましたが、公正取引委員会に意見を提出したい人の参考をつくりました。 docs.google.com/document/d/1... 次の質問に回答すると、それが今回のパブコメでの素晴らしい意見になります。また、意見を提出しないまでも、ざっと一読しておいていただくだけでも嬉しいです。 (質問1) データ保有者(創作者、クリエーター)は、生成AIモデル開発事業者やユーザーのどのような行動が、公正でないと考えていますか? (質問2) 国内の事業者や生成AIユーザーは、学習データの取得に慎重なのでしょうか?

公正取引委員会の生成AIのパブリックコメント、22日の18:00までです。 終日ではないので明日が描ける最終日と思っていいと思います。 何度も書いている事ですが、 現状に疑問を持っているけど表では様々な事情があって話題に触れられない…という方は特に、こういう行動出来る機会を大事にして欲しいと思います。 どんな考えがあってもこういった機会に何も出来なければ居ない事と同じにされてしまいますし、この問題に関しては、誰かが自分の代わりに全てを解決してはくれません。

There's something I need people to understand. The studio that made this released some numbers. They had to produce 85 minutes of footage to find 15 usable seconds for this ad. The machine that creates moving images out of nothing has a failure rate of 340:1 and the best 15 seconds... is this.

gigazine.net/news/2024112... ↓ 著作者に対するAIデータ使用料が「3年1冊にあたり38万円」 という情報。 これは妥当なのでしょうか。いったんAIデータに利用されれば永久に切り刻まれた著作物は拡散されるのでは?拡散されたものを3年で終わりだからAIデータから削除してと申し出てすでに拡散された著作物はどうなるのか。 ちょっと具体的なメカニズムが理解できていません😅

artists 🎨 stop apologizing for not posting, for taking breaks, for being slow, for drawing the same character, for posting only sketches, for gushing over OCs, for being a beginner, for raising your prices, for trying a new medium— start doing whatever you heckin want

11月22日18:00必着で、公正取引委員会が生成AIについての意見募集をしています。 見るのが嫌、考えたくない、って人も居るかもしれませんが、それでも必要な事だと思うので共有しておきます。 普段SNS等で意見が言えないって人ほど、こういう意思表示の貴重な機会を大事にして欲しいです。 www.jftc.go.jp/houdou/press...

#art