The hottest AI Art Substack posts right now

And their main takeaways
Category
Top Art & Illustration Topics
Astral Codex Ten β€’ 25259 implied HN points β€’ 20 Nov 24
  1. Many people struggle to tell the difference between AI-generated art and human-created art. Most scored just slightly above random guessing on a test designed to distinguish between the two.
  2. Participants often judged art based on its style rather than its origin. People tended to think that certain styles, like Impressionism, had to be human even when they were AI-created.
  3. Surprisingly, more people preferred AI art over human art, with many claiming they disliked AI art but still choosing AI pieces as their favorites.
The Rectangle β€’ 56 implied HN points β€’ 06 Mar 26
  1. AI-generated videos have become visually convincing, and that has encouraged lots of people with little originality to churn out derivative, often copyright-leaning clips like endless fight scenes.
  2. Posting these AI videos feels self-centred and unengaging for viewers β€” it's like hearing someone describe their dreams, with no real creativity or dialogue invited.
  3. AI can help filmmaking as a tool (crowds, sets, effects), but fully AI-made movies will likely lack the storytelling depth and emotional connection that make art worthwhile.
Astral Codex Ten β€’ 10048 implied HN points β€’ 14 Oct 24
  1. There's an AI Art Turing Test happening, which is a fun way to see if you can tell the difference between AI-generated art and human-made art.
  2. Participants can take part by filling out a form, which takes about twenty minutes, and the results will be shared the following week.
  3. Don't peek at the comments until you're done with the test, as they might give away answers.
The Algorithmic Bridge β€’ 456 implied HN points β€’ 13 Nov 25
  1. Ghosts are a part of our culture and represent our fears of the unknown. They're not just stories from the past; they symbolize deeper issues we face today.
  2. AI can evoke strong emotions and help people express feelings they struggle to communicate. It can be a tool for creativity, even in times of grief, as seen in stories written with AI assistance.
  3. Using AI in art and literature raises questions about creativity and originality. There’s a debate about whether AI can truly create art or if it simply reshuffles existing ideas and concepts.
The Common Reader β€’ 1382 implied HN points β€’ 07 Jul 25
  1. AI art could become popular if it keeps getting better. People have enjoyed machine-made art before, so there's a chance they’ll like it now.
  2. Art has changed with technology many times before. Just like past mechanical art, AI art might also find a place in people's hearts.
  3. People might value AI art for different reasons, like connecting with emotions or memories. If it becomes unique to each person, it can still feel special.
Get a weekly roundup of the best Substack posts, by hacker news affinity:
So Here’s a Thing β€’ 1573 implied HN points β€’ 28 Feb 23
  1. The featured AI-tinged images in the announced Substack post provide a glimpse into the world of pastel retro-futurism art on social media.
  2. AI art sparks a discussion similar to digital art in the 1990s, where it can be a tool for artists to bring visions to life rather than an entity of great intrinsic value.
  3. The art pieces evoke a sense of capturing Ray Bradbury's mind after a night of adventures with David Bowie, creating a unique and vivid aesthetic.
Break Free with Karen Hunt β€’ 1218 implied HN points β€’ 02 Jun 23
  1. Conservatives boycotting companies for promoting the transgender agenda may not be effective as the focus is on influencing the younger Generation Z.
  2. Boycotting a few companies may not be enough to stop the transgender agenda that is progressing towards a future where AI and technology dominate.
  3. Young people are increasingly embracing fluid gender identities and diverse pronouns, reflecting the evolving societal norms.
Breaking Smart β€’ 74 implied HN points β€’ 15 Dec 25
  1. A simple doodle evolved into a 41-piece iPad painting series called Bucket Art that uses the flood-fill (bucket) tool and pareidolia to discover motifs like waterfalls and ships.
  2. The hand-painted set was used to fine-tune a generative model on titles.xyz, letting many people create, mash up, and publish new Bucket Art images derived from the style.
  3. The platform links training images as NFTs to provide clear provenance and micropayment flows to contributing artists, enabling cheap, auditable, high-volume generative art and new creative-composability dynamics.
Counter Craft β€’ 421 implied HN points β€’ 08 Jun 25
  1. Using AI in art can be great if it's done thoughtfully. It's important to be open about how AI is used and ensure it enhances rather than replaces creativity.
  2. There are successful examples of authors using AI to help craft dialogue or ideas, but they are intentional and clear about their AI use. This distinction separates meaningful art from low-quality outputs.
  3. For artistic integrity, it's best to be honest about your use of AI, especially if it generates a lot of the work. Just like with plagiarism, taking someone else's work, including AI-generated text, without credit is not acceptable.
Daniel Pinchbeck’s Newsletter β€’ 3 implied HN points β€’ 28 Feb 26
  1. A hands-on creative studio offers weekly live sessions, mentorship, and an active community to learn AI-native and hybrid storytelling. Membership is paid and recordings are available to help people catch up.
  2. Using AI for video is powerful but quirky: it can produce spectacular effects yet struggles with simple things like consistent character dialogue. Working with it is an artisanal, iterative craft as tools evolve rapidly.
  3. The effort focuses on shaping ethical, hybrid approaches that humanize AI and protect creative roles, while exploring business and distribution strategies to counter low-quality AI content. The aim is to expand cinematic language without replacing actors and technicians.
Cybernetic Forests β€’ 279 implied HN points β€’ 04 Feb 24
  1. Art can either sell systems of power or support technocracy, highlighting the importance of using art to question and challenge technology rather than serve as a tool for propping up power.
  2. Imagination is often linked with AI, but it's crucial to move beyond speculative thinking to consider the real impacts and consequences of AI on our world today.
  3. Artists, when working with and against technology, can play a role in subverting and challenging powerful systems by acting as parasites, critiquing and revealing flaws instead of just showcasing technological prowess.
Cybernetic Forests β€’ 199 implied HN points β€’ 21 Jan 24
  1. When creating images with AI, we are essentially building data visualizations based on training data, and this can lead to reproducing stereotypes found in the training data.
  2. Archives, like Wikimedia Commons, require curation and community engagement to ensure responsible and equitable representation in AI training datasets.
  3. There is a need to recognize the cultural and emotional value of images and data, and to approach AI training data as more than just facts, but as part of a larger social and cultural fabric.
Things I Think Are Awesome β€’ 196 implied HN points β€’ 31 Dec 23
  1. The post discusses the Welsh tradition of Mari Lwyd, a skeletal horse pranking people during New Year's Eve.
  2. Mari Lwyd is associated with village mummers in south Wales in the 1800s and is constructed of a horse's skull and a white sheet.
  3. The post also covers various AI art tools, including an AI animation artist called NiceAunties and AI-generated SVG from images.
Cybernetic Forests β€’ 239 implied HN points β€’ 27 Aug 23
  1. Generative AI and Donald Trump's mugshot show how media signals and internet metrics influence image generation, highlighting the impact of feedback loops on visual media.
  2. Midjourney's prompts reveal how AI interprets and categorizes images based on stylistic markers like Sumatraism, showcasing a blend of aesthetic information and stereotypes.
  3. Algorithmic Hauntology warns about the dangers of automated pattern analysis rooted in biased archives, emphasizing the need to approach AI predictions and data processing with caution and awareness of historical influences.
RUINS β€’ 137 implied HN points β€’ 12 Jan 24
  1. AI art generators are advancing rapidly, with the ability to interpret text prompts and produce diverse imagery.
  2. There is a contrast between AI-generated art and human art in terms of process, risks, and meaning.
  3. The role of AI in art creation is debated, questioning whether it should supplement or replace human creativity.
Bet On It β€’ 90 implied HN points β€’ 12 Aug 25
  1. There's a contest for AI artists to create unique game chips. If you impress with your work, you could win up to $500.
  2. The challenge is to beat a specific set of art created by a human artist named Savane, who is known for high-quality work.
  3. Artists need to submit their artwork by September 15, and the winner will be decided by personal preference, so it's a tough competition.
Cybernetic Forests β€’ 119 implied HN points β€’ 31 Dec 23
  1. Subscribers to the newsletter tripled this year, showing growth and increased interest in the content.
  2. The author created and taught a course on AI Images, which was referenced by other educational institutions, showcasing influence in the field.
  3. The Algorithmic Resistance Research Group presented at the DEFCON 31 AI Village, demonstrating involvement in cutting-edge AI art and research.
Things I Think Are Awesome β€’ 176 implied HN points β€’ 01 Aug 23
  1. AI-run haunted houses in literature often present intriguing narratives and questions about space and identity.
  2. Poets and writers can explore unreal architectures and dream-like spaces created by AI tools to tell unique and engaging stories.
  3. Exploring the intersection of AI, art, and poetry can lead to the creation of cosmic narratives within the digital labyrinth.
The Algorithmic Bridge β€’ 265 implied HN points β€’ 27 Nov 24
  1. Art has two layers: a visible surface like colors and shapes, and a hidden layer that includes history and culture. AI art usually lacks this deeper meaning.
  2. People often struggle to tell AI art from human-made art because they focus only on the surface. They can learn to spot AI art by asking if it has that deeper history and consistency.
  3. Human creativity is stronger because it connects to real experiences and truths. AI can mimic but it doesn't understand the world or the meaning behind art.
Cybernetic Forests β€’ 159 implied HN points β€’ 23 Apr 23
  1. The story 'Sarah Palin Forever' explores the impact of political environments and media ecosystems on shaping identities.
  2. The concept of deepfakes and how they can be used as tools for satire and storytelling is discussed.
  3. The power of images and words in media is highlighted, emphasizing how narratives shape our perceptions and understandings.
Things I Think Are Awesome β€’ 157 implied HN points β€’ 14 May 23
  1. The project combines old haiku poetry with new AI technology to create multimedia experiences.
  2. The haiku illustrations include AI-generated letters, music, and video clips, focusing on the atmosphere and emotions of the poems.
  3. The project showcases the challenges and creativity involved in using AI models to generate visuals and music based on poetry prompts.
Cybernetic Forests β€’ 119 implied HN points β€’ 26 Nov 23
  1. Generative Adversarial Networks (GANs) were innovative in AI art, generating images based on existing datasets and patterns.
  2. Artists using GANs had more control over their datasets, shaping the outputs with their own images and deciding what to include, unlike modern Diffusion models.
  3. Training and working with GANs was an experimental process, where artists had to understand the algorithm's perspective and engage in a technical dialogue to create art.
Wood From Eden β€’ 528 implied HN points β€’ 18 Mar 24
  1. AI-generated images are becoming prevalent on the internet, but their superficiality can hinder deeper meaning and understanding.
  2. The proliferation of AI images can crowd out images of real people and places, which traditionally provide more valuable information and opportunities for discovery.
  3. In a digital age where sharing reality is crucial, efforts should be made to prioritize genuine, real-world images over AI-generated ones to maintain authenticity and meaning.
ARTβ‹‚CODE β€’ 16 implied HN points β€’ 04 Dec 25
  1. A generative model trained on many personal photos can show an emergent, proto-perception that feels similar to natural intelligence. Its glitches and odd outputs reveal how representations form.
  2. Keeping and animating the model’s discarded glitches and textures can produce slow, hypnotic audiovisual landscapes that make the medium’s limits and character visible. These pieces are being shared as work-in-progress performances rather than immediate online releases.
  3. The project asks us to re-see recent technological and artistic advances amid cultural fatigue, political distrust, and loneliness, warning that mass-produced, shallow content can desensitize us. It calls for renewed intention and appreciation of how remarkable the past few years have been.
Cybernetic Forests β€’ 139 implied HN points β€’ 13 Aug 23
  1. The Algorithmic Resistance Research Group (ARRG!) focuses on critiquing and analyzing AI systems, highlighting issues like data rights, stereotypes in AI output, ecological harms, political risks, and the impact of red teaming.
  2. ARPG! highlights the importance of challenging the logic of AI systems to avoid exploiting stereotypes, artist data rights, and push back against automated cultural production.
  3. Research showcased the use of Gaussian Noise Diffusion Loop to create abstract art, challenge content moderation tools, and explore the dynamics of AI-generated imagery.
Cybernetic Forests β€’ 119 implied HN points β€’ 22 Oct 23
  1. Refik Anadol's 'Unsupervised' art at the MoMA uses AI to visualize the MoMA's art archives in a unique way, offering a new perspective on data analytics and art
  2. Anadol's art piece juxtaposes the complexity and mystique of AI systems with the potential for human understanding and engagement, sparking discussions on the implications of AI in art and society
  3. Alternative models of AI art, like 'Anatomy of an AI System' and 'What Models Make Worlds', present critical perspectives that question the power dynamics and ethical implications of AI, contrasting with the awe-inspiring presentation of AI in Anadol's work
Cybernetic Forests β€’ 119 implied HN points β€’ 30 Apr 23
  1. Human perception of images is deeply intertwined with personal experiences and emotions, shaping how images are interpreted and associated with memories.
  2. Creating art involves a fusion of individual lived experiences and learned skills over time, contrasting with the quick generation of images by AI devoid of personal experiences.
  3. AI images are structured based on categories and datasets, emphasizing the need for artists to negotiate these categories and infuse individualized interpretations into the process.
Litverse β€’ 79 implied HN points β€’ 09 Dec 23
  1. The process of creating art with AI can be a mix of excitement and uncertainty, similar to Frankenstein's creation of his monster.
  2. Artists are facing a new age of AI technology and must adapt to the changes it brings, even if it sparks fear and ethical concerns.
  3. Utilizing AI in art could be a way to reach more people and make impactful creations, but it also presents challenges in maintaining human connection and relevance.
Do Not Research β€’ 99 implied HN points β€’ 25 Oct 23
  1. Explore the intersection of AI technology and human society through Daniel Felstead & Jenn Leung's project 'Literally No Place' on Dis.art.
  2. Consider the diverse range of futures AI technology could bring, from a post-scarcity society to potential human extinction.
  3. Reflect on the different perspectives presented in the project before forming a stance on AI's impact on society.
Cybernetic Forests β€’ 139 implied HN points β€’ 18 Dec 22
  1. Reflection on the problems and implications of AI-based image generation in art
  2. Consideration of the origin and context of AI training data, highlighting issues like exploitation and biases
  3. Exploration of rethinking AI images as material for artistic expression, and the importance of artists reclaiming agency over these tools and the images they create
Cybernetic Forests β€’ 59 implied HN points β€’ 16 Aug 23
  1. The author is offering a special promotion for readers to support their newsletter with a $1.00 monthly donation until September 1st.
  2. The newsletter includes interviews with new media artists, explorations of generative art practices, examinations of generative AI, and historicizing cybernetic and generative art work.
  3. Readers are encouraged to support the newsletter with financial contributions or by sharing it with others who might appreciate its content.
Cybernetic Forests β€’ 79 implied HN points β€’ 12 Feb 23
  1. Diffusion models start by generating random noise and work backward to create images based on prompts.
  2. The model aims to remove noise based on the prompt, creating a recursive process of noise refinement.
  3. Diffusion models struggle with abstract details like Gaussian noise, leading to errors in representation.
Generative Arts Collective β€’ 79 implied HN points β€’ 22 Dec 24
  1. You can create unique snowflake designs using a website that generates them. It's fun to play around with the designs and save your favorites as SVG files.
  2. Generative AI is being used in interactive art to create experiences that change based on audience interaction. This can highlight important issues like environmental damage.
  3. Using voice commands to control software like Blender shows how technology can make creative tools more accessible, allowing for new and exciting ways to interact with art and design.
Cybernetic Forests β€’ 59 implied HN points β€’ 29 Jan 23
  1. Refik Anadol's AI art piece, 'Unsupervised,' at MoMA uses AI to interpret and reimagine the history of modern art, creating a mesh of pixelated visuals.
  2. Interpolation in AI refers to filling in the gaps between data points or images, creating a smooth transition and possible new variations.
  3. The concept of interpolation extends to creating a connection and kinship between disparate entities in an artistic representation, showcasing the latent possibilities in the in-between spaces.
Marginally Compelling β€’ 24 implied HN points β€’ 29 May 25
  1. AI can be a helpful tool for speeding up creative projects and making art, but it also raises big questions about our future.
  2. One important AI art piece is a completion of Keith Haring's unfinished work, which reveals deeper messages about art and technology.
  3. This artwork makes us think about our relationship with AI and what we value in art and creativity.
Cybernetic Forests β€’ 39 implied HN points β€’ 19 Feb 23
  1. Eryk Salvaggio will be a guest on BBC 4's _Digital Humans_ radio show discussing creativity and automation in AI art
  2. Sensitive Noise is an AI-generated artwork exploring content moderation and human sensuality through censored images
  3. Apply for STORY x CODE, a residency program focusing on AI and Machine Learning in storytelling, filmmaking, and animation
Crypto Good β€’ 3 implied HN points β€’ 05 Dec 25
  1. Many people don't read long articles anymore; they skim through them for quick information. Turning writings into comic strips makes the message more engaging and memorable.
  2. You don't need to be a skilled artist to create cartoons anymore. AI tools can help anyone generate comic strips quickly and easily, no prior experience required.
  3. Creating comics with AI only takes a little conversation about style and mood, and then you can transform your ideas into fun, visual formats that people enjoy.
Critic at Large β€’ 19 implied HN points β€’ 15 Aug 23
  1. The MoMA displayed an artwork called "Unsupervised" by artist Refik Anadol, which uses AI to create constantly changing images based on data from 180,000 artworks and real-time external factors.
  2. Anadol's AI artwork, while technically impressive, lacks the depth and connection to traditional art that keeps viewers engaged and leaves them feeling unfulfilled.
  3. The comparison between Anadol's AI artwork and a traditional piece like Rothko's painting highlights the importance of human creativity and emotion in art that AI may struggle to replicate.