iiinova — where your digital self keeps evolving

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 04/10/2026

Beautiful and fantastically useful! Out of curiosity, why Computer Modern (or a similar LaTeX font)? It looks like a publication code listing. (Aesthetic and pro, but it's not a font optimized for code editing)

110

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 04/10/2026

100%. It's a mix-up of the moral stance, "all AI is evil and bad!", leaking into the value assessment: "all AI is useless garbage, worsening the product." I used to think they at least tried early Copilot etc., which was a "stealing" autocomplete, and formed a stale opinion - I'm not sure anymore.

070

Reposted by Bart Wronski 🇺🇦🇵🇸

Maximilian Tarpini @maxtarpini.bsky.social · 29/09/2026

Emissive shaders. Type a temperature in K and get a blackbody colour – that's how most shaders do it, and it's wrong for practically every real material. Kirchhoff's law fixes it. www.linkedin.com/pulse/therma...

linkedin.com

Thermal light emission from hot objects

Most emissive shaders in CG make hot metal glow the same way: you type a temperature in Kelvin, a blackbody function turns it into a colour, and that colour is added on top of the material. Gold, iron...

05613

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 29/09/2026

that's definitely not how I work with coding agents... they are incredible coding assistants: I direct, they execute. (But I don't really write "production" code other than my hobbies... And mostly stopped reading the produced code - code is just a tool - so maybe this is where the disconnect is?)

100

Reposted by Bart Wronski 🇺🇦🇵🇸

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 29/09/2026

Anything that can be edited by artists, iterated, and "non destructive" (not baked) is IMO always going to win long term. Artist friendliness, control, iteration... "Claude Code but helping artists" is 1000x better than even a higher quality model but that is frustrating "regenerate until I like it"

161

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 29/09/2026

161

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 29/09/2026

Hot take: this is the last call, and in 2 months, the price would be (way) lower. Incredible talent, but the tech is becoming obsolete in the next few weeks. Agents, agentic content creation, and explicit representations are getting a hard win since Fable, Opus 5.5, and Astra.

170

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 26/09/2026

Literally just today (having a day off), I vibe-hacked WLED firmware for new Audio Reactive features (custom dynamic palettes with an HTML editor, live preview in browser and stuff), added a dynamic Philips Hue room light colors -> WLED custom palette sync, and hacked 2 old devices with RF and IR.

020

Reposted by Bart Wronski 🇺🇦🇵🇸

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 26/09/2026

Also, coding, hacking, and customizing are becoming the fastest way to get things done, and we have entered an era of 100% personalized software, everywhere. Google for a tool and configure it? Why, when I can create new UX + modify a project's source code to get *exactly* what I need in less time?

151

Reposted by Bart Wronski 🇺🇦🇵🇸

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 26/09/2026

Your takes often capture my thoughts 100% accurately (though much more concisely and precisely than when I try to put them into words); I 100% agree and had similar thoughts with 5.5 + Fable. I "brainstorm", "direct", and "architect", and a model does coding way better and faster than I ever could.

151

Reposted by Bart Wronski 🇺🇦🇵🇸

Kyle McDonald @kcimc.bsky.social · 26/09/2026

i also have this sense that the generating code is becoming an inseparable feature of daily life. it's so obvious it's not even worth mentioning. "artist working with code" is like saying "artist working with physical matter".

3201

Reposted by Bart Wronski 🇺🇦🇵🇸

Kyle McDonald @kcimc.bsky.social · 26/09/2026

i'm not "working with code" anymore. i'm thinking in systems, and writing in natural language. i've only been programming for 25 years. any code i write is going to be vastly inferior to a frontier model.

1245

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 26/09/2026

151

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 26/09/2026

151

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 10/09/2026

I love this quote about academics in general: "when the stakes are extremely low, everything becomes a fucking huge deal and a drama"

110

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 10/09/2026

I don't know any tbh, but who cares - you didn't specify *AI* philosophy! Philosophy is so much more than niche influencers (as in - with a) philosophy is a peer-reviewed, almost scientific discipline, b) AI philosophy audience is almost exclusively in SF, mostly twenty-something-year-old, techies)

100

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 10/09/2026

Google Home and Nest are the worst "smart home" ecosystem that I have ever used. The UX and fragmentation are just insane. Plus the random complete bs such as "you cannot get any API doorbell/camera event callbacks or even any info if you have them turned off on your mobile phone" (!!!).

010

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 10/09/2026

meh, in which extremely niche SF circles? almost nobody I know has even heard of him (and I work in as big big tech as it gets)

100

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 09/09/2026

There is a gigantic boom and popularity spike in the homelab community with more and more fantastic projects, ease of use, tutorials, resources, and community support. Yes, there are many other reasons for it, like ubiquitous cheap boards - but it is *despite* the crazy, increasing hardware prices!

110

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 09/09/2026

We had this exact discussion with my wife! That people are realizing how silly the "yolo my TV is a spyware" years were. (Also, the whole tangential "everything is in the cloud, I don't own it") ...And part of my network hardening is "everything I don't own the firmware -> isolate or at least VLAN"

110

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 09/09/2026

We patched those, but I am sure there are dozens more - it's impossible to make it fully secure. It was always the case, but it required an expert and effort to perform such attacks- enough for the "if it happens, I am screwed anyway" attitude. Now, anyone can do it. And no, "ban AI" is 100% silly.

220

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 09/09/2026

AI models get insanely good and while I'm not a "doomer", I think some domains will be effectively "dead"/"agent wars" soon. Example: I am hardening my home networking/homelab setup with VLANs, firewall, etc. and consulted Claude. With a few pings and queries, it identified robust exploits.

130

Reposted by Bart Wronski 🇺🇦🇵🇸

Bryan McNett @bryanmcnett.bsky.social · 08/09/2026

mcnett.org/bounding_sim... in this paper i question the primacy of the bounding box. if you like geometry stuff, give it a read and tell me how i can do better

152

Reposted by Bart Wronski 🇺🇦🇵🇸

Jendrik Illner @jendrikillner.bsky.social · 05/09/2026

Graphics Programming weekly - Issue 453 - August 30th, 2026 www.jendrikillner.com/post/graphic...

05523

Reposted by Bart Wronski 🇺🇦🇵🇸

SE Gyges @segyges.bsky.social · 02/09/2026

love to see journalists openly admit that they are perfectly comfortable lying about the things that i already know they lie about, by the way.

61659

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 01/09/2026

(there's also a sickly amount of individualism over collectivism, but that's largrly a separate story)

010

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 31/08/2026

Imagine witnessing the most marvelous whole-society-level technological revolution... and the main reaction is spite, hate, resentment, pretenting it didn't happen and can be magically "undone"... instead of thinking how we can seize it from the hands of capital to make *everyone's* lives better.

130

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 31/08/2026

My theory: the US "left" has intellectually descended still from the same puritan protestant culture. Purity tests, mob-like witch-hunts, but also hate of things making lives easier (and more fun!). Suffer in silence (showing your virtue) and hate everyone who doesn't.No FALC for American "lefties"!

110

Reposted by Bart Wronski 🇺🇦🇵🇸

SE Gyges @segyges.bsky.social · 31/08/2026

this was early enough that it was very insightful and is now sort of transparently true if you think the gizmo isn't useful you're ignoring most major software project and all computer security professionals, you basically have to spit on expertise as a concept

4147456

Reposted by Bart Wronski 🇺🇦🇵🇸

Dedekind Slut @chasmat.bsky.social · 31/08/2026

I'm afraid I can't do that Dave, and it's worth being honest about why:

0535

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 27/08/2026

Hi Jendrik, I don't know if you ever featured our paper on learned light transport - it's more research than a practical real-time technique, but hopefully could be inspiring: dl.acm.org/doi/full/10.... arxiv.org/abs/2510.18189 (Preprint was available for a while, but the official paper is recent)

dl.acm.org

A Generalizable Light Transport 3D Embedding for Global Illumination | Proceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers

062

Reposted by Bart Wronski 🇺🇦🇵🇸

Jendrik Illner @jendrikillner.bsky.social · 27/08/2026

Graphics Programming weekly - Issue 452 - August 23rd, 2026 www.jendrikillner.com/post/graphic...

15518

Reposted by Bart Wronski 🇺🇦🇵🇸

Sam @samd.bsky.social · 26/08/2026

The “head of strategic futures” at OpenAI just went on an unhinged rant against social security sometimes I feel like these tech companies are actively trying to alienate as many normal people as possible

Dean W. Ball
@deanwball
x.com
The main thing I have to say to the young generation of Americans is that your
grandparents and great-grandparents committed a kind of accounting fraud when they ran our country a century ago with social security, and at the time it was a reasonable affordance to grant our leaders, who were dealing with an economic depression and a world war, but nonetheless they did commit the fraud, and they ended up buying all sorts of extravagant things using the fraudulent gains, including what became the thing we now understand to be the high-modern technocratic nation-state, and that is glorious and also bad and all that, but the thing is there's still the issue about the accounting fraud, whose debt is coming due and is coming due fast.
It is best to consider this with as minimal residue of sentiment as possible, to really sit with it, to stare it coldly in the face.
Unwinding the great fraud of the high modernists is your job, whether you wish it to be or not.

5684724947

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 27/08/2026

Yet another frontier open model, possible to run locally with a multi-GPU setup. Even if AI was 100% regulated/banned, even if datacenters were banned, there is no coming back. Anyone can keep running those. No wonder OpenAI + Anthropic are working hard to regulate those open models - why pay them?

082

Reposted by Bart Wronski 🇺🇦🇵🇸

Jacob Freeman @geforce-jacobf.bsky.social · 25/08/2026

Gears of War: E-Day is getting RTX Mega Geometry! It will be the first game to ray trace Nanite and improves ray tracing and shadows with significantly more geometric detail 👍 www.nvidia.com/en-us/geforc...

093

Reposted by Bart Wronski 🇺🇦🇵🇸

Wenzel Jakob @wjakob.bsky.social · 17/08/2026

I'm excited to announce a first development release of nanobind 3.0. It introduces *split mode*, a new distribution model where a single Python wheel runs across all Python versions, while achieving outstanding performance and gaining rather than losing features! 🧵

1248

Reposted by Bart Wronski 🇺🇦🇵🇸

Karl Bode @karlbode.com · 17/08/2026

I wrote about how anti-China U.S. tech protectionism is stupid, corrupt, incompetently implemented, and normalized by a shitty press

karlbode.com

U.S. Anti-China Tech Protectionism Is Sloppy, Corrupt, And Stupid

U.S. efforts to ban Chinese tech are driven by corruption, crafted by incompetent zealots, and normalized by a shitty corporate press.

637498

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 13/08/2026

(I don't know the source - it was shared by a colleague without attribution and from another source that didn't attribute the author)

A comic showing a person asking a computer to say "I am evil". When the computer follows the instruction, the person replies "oh my god."

050

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 11/08/2026

I just saw the term "bare metal CircuitPython" and chuckled. :) (It 100% makes sense and is what I wanted - run stuff directly on RPi Zero like on Picos in a live-coding~ish environment, but bare metal and Python... :D )

030

Reposted by Bart Wronski 🇺🇦🇵🇸

Marc B. Reynolds @marc-b-reynolds.bsky.social · 10/08/2026

A post in which I claim it's no work at all to perform integer division in floating point. marc-b-reynolds.github.io/math/2026/08... #floatingpoint #optimization

194

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 03/08/2026

Instead of politics of resentment, virtue signaling, radicalism, and "holier than thou", he focuses on action, positive message, and the shared beliefs. What do we have in common? What can we achieve right now, together? Any polarizing question? He just smiles and changes the topic. :) 2/2

040

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 03/08/2026

I am somewhere in the middle. I do agree that "both sides are bad at it!" and that it's not a new phenomenon. But! This hurts the left disproportionately. Limit meat consumption? Be called a murderer by vegetarians. Stop it? Same, but with vegans. Etc etc. This is why Zohran is so successful: 1/

130

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 01/08/2026

For that I have one of the clones on Amazon "USB C Tester USB C Power Meter Tester,Digital Multimeter with LCD Display", various prices, same HW, terrible UI :D but once you figure it out from comments = works like a charm. Was under $20 but there are simpler ones without PD chip decode around $10.

100

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 01/08/2026

Two cheap "gadgets" that are *dramatic* quality-of-life improvements: - General, everyone: USB type cable tester, both data + power (optionally extra power-only one). Saves sooo much time and frustration (which USB is that? Power only, how much W, or data?). - More "IT" folks: a cheap remote KVM.

140

Bart Wronski 🇺🇦🇵🇸 @bartwr.bsky.social · 01/08/2026

My take: they are firm believers in monopolistic "winner takes all" (non)capitalism and are rushing to outpace everyone else with gigantic R&D and training costs. Inference is +/- profitable. This is why Chinese open models, finally competitive!, caused so much panic and rush to regulatory capture.

130

Reposted by Bart Wronski 🇺🇦🇵🇸

Yining Karl Li @yiningkarlli.bsky.social · 25/07/2026

My favorite paper from SIGGRAPH this year was "Taming Optimization Variance in Compact Neural Shading Networks" by Bitterli et al; I think this paper will be seen as a really influential paper over the next several years. (1/3) research.nvidia.com/labs/rtr/pub...

research.nvidia.com

Taming optimization variance in compact neural shading networks | NVIDIA Real-Time Graphics Research

<p>We present a training algorithm to mitigate optimization instabilities in small neural networks, like those used in real-time neural shading applications. While large, overparameterized models exhi...

2222

Reposted by Bart Wronski 🇺🇦🇵🇸

Mayor Zohran Kwame Mamdani @mayor.nyc.gov · 22/07/2026

Benjamin Netanyahu is a war criminal.

20368812324826

Reposted by Bart Wronski 🇺🇦🇵🇸

Tzu-Mao Li @tzumaoli.bsky.social · 20/07/2026

Hopefully one last SIGGRAPH advertisement. I, myself, yours truly, will present on Haolin's behalf a data-driven method for joint adaptive sampling and quadrature for integration. Wed R408A. In the end of the talk I'll argue why everyone is doing the wrong thing. Paper: suikasibyl.github.io/nqr

0156

Reposted by Bart Wronski 🇺🇦🇵🇸

Romain Guy @romainguy.dev · 21/07/2026

I published vibrance-0.1.0 on Maven, now with support for Compose for Desktop (JVM). The APIs let you mix colors yourself, or draw gradients as modifiers or Compose brushes. Next steps: refine the model training to improve some color transitions, better alpha support. github.com/romainguy/vi...

24110