Wrote up the full architecture of Skiagrafia, a local-only batch image-to-SVG pipeline for Apple Silicon.
tsevis.com/every-pixel-is-a-tesserae
github.com/tsevis/skiagrafia

I built it for creative people: designers, illustrators, artists, mosaic makers, educators, cartography nerds, poster makers, and anyone who wants a map not as a screenshot, but as material to transform.
github.com/tsevis/Hippa...

I made a new open-source app called Hipparchus.
It creates clean vector maps from OpenStreetMap data and exports them as layered SVG files. That means roads, buildings, water, parks, labels, railways, coastlines, and other elements stay separate and editable.

It uses local, open-source vision-language, computer vision, and language models to segment content, refine masks, and produce structured outputs for design and production pipelines.

Dear friends, I just released Skiagrafia: a desktop app for batch-generating masks and vector silhouettes from folders of images.
Repo: github.com/tsevis/skiag...

Charis Tsevis - Beyond “Vibe Coding” Visual designer living on the sunny side of the planet. Photomosaics, computergraphics, neofuturism, digital art and graphic design

For more than two years now, I’ve been experimenting with building software tools with AI assistants.
People often call it #vibecoding. Personally, I never liked the term.
Coding is only a language. The real goal is building systems that do something useful.
tsevis.com/vibecoding-o...

GitHub - tsevis/MacWhisperer: MacWhisperer is a macOS desktop GUI for OpenAI Whisper that makes local speech transcription and English translation easy, with auto model loading, live recording, input ... MacWhisperer is a macOS desktop GUI for OpenAI Whisper that makes local speech transcription and English translation easy, with auto model loading, live recording, input device controls, and file-b...

MacWhisperer: Private, local Whisper for Mac. Turn media into editable text. github.com/tsevis/MacWh...

AI Coding Plan-Code Freely. Ship Faster. No Surprise Bills. - Alibaba Cloud Experience top-tier performance at a fixed monthly price. Qwen3-Coder-Plus, the cost-effective choice. Works with Cline, Claude Code & Qwen Code.

As a long-time Qwen Code user who has also explored MiniMax, Kimi, and GLM, I'm eager to evaluate how this integration optimizes my development workflow.
www.alibabacloud.com/en/campaign/...

I've officially upgraded to the Qwen Coding Plan via Alibaba Cloud. While the initial setup required a bit more effort than anticipated, the value proposition is undeniable: access to 8 models for just $10/month. #AICoding

Bruno Munari and the quest for comfort in an uncomfortable chair 25 years without the tutelary deity of good design, an all-round artist in exploring all fields of creation and expression. Including surreal irony, such as that of his lucid provocation targeting the...

Munari’s call to action is to work for Man, not just for "creativity (or bizarreness)." A timeless reminder that true innovation can also mean refinement, and that placing a simple, functional object in the right place can be a radical act.
www.domusweb.it/en/from-the-...

Why do we obsess over inventing a "never-before-seen" chair, instead of perfecting the most comfortable one we already have? He uses wit to critique design’s endless pursuit of novelty, those "floral hammers" and "baroque" furniture pieces that so often sacrifice an object's true purpose.

I was looking through some old Bruno Munari material and stumbled upon his brilliant 1944 manifesto, "Seeking Comfort in an Uncomfortable Chair." In it, he asks a question that stopped me:

So I choose to do that through my art.
Criticising and fighting are easy.
Proposing, supporting, and building are much harder.
This is my journey in politics, for anyone who wishes to read it.

Charis Tsevis - My Art for Politics Visual designer living on the sunny side of the planet. Photomosaics, computergraphics, neofuturism, digital art and graphic design

People like to put labels on you. Especially in politics.
For me, it has always been difficult to fit into a simple political category.
And yet, as someone who believes in democracy, I also believe in participation and in taking a stand.`
tsevis.com/my-art-for-p...

RIP Giorgos Papadakis
1951-2016

Digital mosaic created with hundreds of photos from his 44 years at Ant1 tv Greece. Commissioned by the broadcaster last year as a gift to him at his last show.

While the benchmarks (HumanEval, MBPP) look impressive, maybe a bit optimistic, what’s really interesting is how this model is stepping up to challenge the current frontier models. Don't cancel your Claude Code subscription yet. But try this beauty for free! iquestlab.github.io
#aicoding #vibecoding

The brand new OPEN SOURCE and FREE iQuest Coder is really impressive! In just 80Gb this is a surprisingly capable new coding model trained in three distinct phases. Built through pretraining, instruction fine-tuning, and rigorous alignment, it shows what focused, phased training can do #codingmodel

Most importantly, he proposes thoughtful ideas for a credible path forward.
In this era of polarizing hype and hate, we need more clear-eyed thinkers like Professor Marcus.
www.youtube.com/watch?v=O0H1...

In this recent MIT FutureTech lecture, he narrates the decades-long history of thinking machines, traces the various illusions that have captivated the AI world, and illustrates - with concrete examples - the serious limitations, and even the fundamental "stupidity," of current LLMs and GenAI models

Gary Marcus has long been one of the most courageous critics of GenAI, even before it was mainstream. He is not a hater - he uses AI as a compass - but he does not stand with the hype sellers. #AIBubble

For all of us who love technology and the challenge of human-machine synergy, it's essential to stay grounded. AGI and super-intelligence can serve as a North Star, but they must not become a broken promise. #AI #GaryMarcus

Gemini 3.0 Flash Designer: Easiest & Cheapest Way TO MAKE AWESOME Frontends! YouTube video by AICodeKing

An experiment showing Claude Skills work with Gemini Flash, useful far beyond just Anthropic's ecosystem.
www.youtube.com/watch?v=wxhS...
Claude Skills Github: github.com/anthropics/s...

Adobe Photoshop Source Code When brothers Thomas and John Knoll began designing and writing an image editing program in the late 1980s, they could not have imagined that they would be adding a word to the dictionary.

The @adobe.com Photoshop 1.0 source code:
computerhistory.org/blog/adobe-p...
#computerhistory #ComputerGraphics #DigitalImaging

Apple's SHARP model generates photorealistic 3D Gaussian reps from a single image.
Free and Open source project.
GitHub: github.com/apple/ml-sharp
Paper: arxiv.org/abs/2512.10685

Charis Tsevis - Creativity in the Era of Machine Intelligence Visual designer living on the sunny side of the planet. Photomosaics, computergraphics, neofuturism, digital art and graphic design

These thoughts developed from a talk I gave at Alexander College in Cyprus. What's your creative compass?
Link: tsevis.com/creativity-i...

The curious ones understand something important. Your compass is your purpose, your reason for creating. The handbook is just methods, and methods keep changing. I just wrote about why this psychological cycle matters more than any AI debate.
#GenAI

The pattern is always identical. Some people freeze and wait for clarity. Others attack the new thing, defending familiar territory. But the most interesting response is transformation, turning that initial fear into curiosity and experimentation.
#DesignEducation

Every creative transition triggers the same ancient response—fear of the unknown. Cameras did this. Computers did this. Now AI.
#AICreativity

Creativity in the Age of Machine Intelligence: A Compass and a Handbook
Tuesday, 16 December at 11:30, at the Alexander College, Larnaka, Cyprus
alexander.ac.cy/creativity-i...

For the Computer Vision aficionados out there: Depth Anything 3 is here! ByteDance's latest model is way faster, more accurate, and more versatile than v2 or Meta's VGGT. A game-changer for computer vision, 3D reconstruction, and AR
Github: github.com/ByteDance-Se... ArXiv: arxiv.org/abs/2511.10647