<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Michal Takac - Software Engineer</title><link>https://michaltakac.com/</link><description>Recent content on Michal Takac - Software Engineer</description><generator>Hugo -- gohugo.io</generator><language>en-us</language><lastBuildDate>Thu, 17 Sep 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://michaltakac.com/index.xml" rel="self" type="application/rss+xml"/><item><title>The creative unit is larger than one person</title><link>https://michaltakac.com/blog/the-creative-unit-is-larger-than-one-person/</link><pubDate>Thu, 17 Sep 2026 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/the-creative-unit-is-larger-than-one-person/</guid><description>We usually describe a creative tool as a scene with two participants: a person asks, a machine responds, and we look at the result.
That leaves out most of what gives creative work its meaning. People learn from each other, borrow techniques, argue about standards, keep old work alive and decide what deserves attention. A tool enters those relationships.
A preprint from September 2, Collective creativity in hybrid societies by Mason Youngblood and colleagues, argues for studying whole populations of interacting people and algorithms.</description></item><item><title>When software can pay</title><link>https://michaltakac.com/blog/when-software-can-pay/</link><pubDate>Thu, 09 Jul 2026 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/when-software-can-pay/</guid><description>When I put GrasShopper aside, one of the things I noticed about the competition was that Perplexity&amp;rsquo;s shopping assistant already had checkout for selected products. Mine only ever recommended things, theirs could also buy them.
That&amp;rsquo;s a bigger step than it looks. A program that finds an answer is one thing. A program that pays for something makes a commitment with somebody&amp;rsquo;s money, and then I want to know who gave it the authority (the same question I had about what an agent is allowed to decide).</description></item><item><title>The cloud has a material footprint</title><link>https://michaltakac.com/blog/the-cloud-has-a-material-footprint/</link><pubDate>Tue, 09 Jun 2026 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/the-cloud-has-a-material-footprint/</guid><description>&amp;ldquo;Cloud&amp;rdquo; is a clever name. It gives computation an image without weight, machines or an address.
Most of the time I&amp;rsquo;m happy with that. I don&amp;rsquo;t need to inspect a power station before I send a file. But once we start making claims about the future of production, the infrastructure has to come back into the picture.
In April 2025 the International Energy Agency published its Energy and AI report. In its base case, global electricity consumption of data centres reaches around 945 TWh in 2030, more than double the 2024 level.</description></item><item><title>GrasShopper: time to move on</title><link>https://michaltakac.com/blog/grasshopper-time-to-move-on/</link><pubDate>Thu, 21 May 2026 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/grasshopper-time-to-move-on/</guid><description>Last summer, I wrote about building GrasShopper, an AI shopping assistant that started with a very simple frustration: why does buying a pair of running shoes sometimes turn into three hours of research?
The idea was to describe what you need, let AI go through the options and reviews, and get a useful buyer&amp;rsquo;s guide. I still think that&amp;rsquo;s a problem worth solving. But looking at what&amp;rsquo;s already available, I don&amp;rsquo;t think GrasShopper needs to be another product solving it.</description></item><item><title>Virtual places have real consequences</title><link>https://michaltakac.com/blog/virtual-places-have-real-consequences/</link><pubDate>Sat, 25 Apr 2026 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/virtual-places-have-real-consequences/</guid><description>Years ago I asked whether we could step inside our software, physically, and see what&amp;rsquo;s going on. I still want that, and these days I&amp;rsquo;m building Graspable, where you describe a scene and an AI agent builds it as a WebXR project you can step into. (I also write about XR and AI on Parallel Mind.) But the more I think about places made of software, the more I see that somebody has to design them with the same care as the physical ones.</description></item><item><title>A provenance record is not a truth machine</title><link>https://michaltakac.com/blog/a-provenance-record-is-not-a-truth-machine/</link><pubDate>Sat, 07 Mar 2026 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/a-provenance-record-is-not-a-truth-machine/</guid><description>A photograph can be genuine and misleading at the same time. It can show a real moment and leave out what happened right before it. And a synthetic image can be clearly labeled and useful, like an architectural render. Where something came from matters, but it doesn&amp;rsquo;t tell you what it means.
This is getting more important as convincing media gets easier to produce. We need better records of where things came from, and we shouldn&amp;rsquo;t ask those records questions they can&amp;rsquo;t answer.</description></item><item><title>An agent needs a boundary</title><link>https://michaltakac.com/blog/an-agent-needs-a-boundary/</link><pubDate>Sun, 08 Feb 2026 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/an-agent-needs-a-boundary/</guid><description>When somebody calls a system autonomous, the first thing I want to know is which decisions it&amp;rsquo;s allowed to make. Can it choose the next step? Change the goal? Spend money? Contact another person? Keep running when nobody&amp;rsquo;s watching?
With GrasShopper this is easy, because the agents only do research. They search, go through reviews, compare prices and write a guide. The buying stays with you.
But searching catalogs, comparing specs, choosing a seller, paying and arranging delivery are different acts with different risks, and they involve different people.</description></item><item><title>Maintenance is a creative act</title><link>https://michaltakac.com/blog/maintenance-is-a-creative-act/</link><pubDate>Mon, 05 Jan 2026 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/maintenance-is-a-creative-act/</guid><description>I&amp;rsquo;ve started going to the gym 7 times (I wrote about it a while ago). Starting was never the problem. The enthusiasm lasted about a week every time, and everything after that week was the hard part.
Projects are the same. A launch is easy to recognize because it has a date, a screenshot, maybe a small celebration. Keeping the thing useful for five years has none of those moments, even though it usually takes more thinking.</description></item><item><title>Access is more than a download</title><link>https://michaltakac.com/blog/access-is-more-than-a-download/</link><pubDate>Tue, 09 Dec 2025 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/access-is-more-than-a-download/</guid><description>A design file can travel around the world in seconds. The ability to make something from it is much slower.
To really use that file you need tools, materials, space, time and enough understanding to recognize a dangerous mistake. Often you also need a person who answers the question the documentation didn&amp;rsquo;t think of. When we say a design is accessible because it can be downloaded, we&amp;rsquo;re counting the easiest part of the trip.</description></item><item><title>When structure becomes ornament</title><link>https://michaltakac.com/blog/when-structure-becomes-ornament/</link><pubDate>Sun, 23 Nov 2025 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/when-structure-becomes-ornament/</guid><description>I like seeing how things work: the ribs under a leaf, the joints of a wooden chair, the cogs in a Lego Technic model that nobody bothered to hide.
That doesn&amp;rsquo;t mean every useful structure is beautiful, or that ornament needs an engineering excuse. But sometimes expression and performance can be developed together, and I find those cases the most interesting.
ETH Zurich&amp;rsquo;s Smart Slab is one of them. It&amp;rsquo;s a ribbed concrete floor made in 2018 with 3D-printed sand formwork, where the structural requirements and the geometry were worked out within the same fabrication process.</description></item><item><title>More ideas can still mean less variety</title><link>https://michaltakac.com/blog/more-ideas-can-still-mean-less-variety/</link><pubDate>Sun, 26 Oct 2025 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/more-ideas-can-still-mean-less-variety/</guid><description>When I was shopping for running shoes, I read 7 &amp;ldquo;TOP 10&amp;rdquo; articles and every one of them recommended something different. That frustration is how GrasShopper started. Lately I think about the other side of it. If everybody asks the same assistant, everybody may get pointed to the same few pairs of shoes. For shoes that&amp;rsquo;s probably fine, but for writing, design or music I&amp;rsquo;m less sure.
Anil Doshi and Oliver Hauser tested this with short stories.</description></item><item><title>A render is a promise</title><link>https://michaltakac.com/blog/a-render-is-a-promise/</link><pubDate>Tue, 16 Sep 2025 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/a-render-is-a-promise/</guid><description>A convincing image of a building comes with many claims that you can&amp;rsquo;t see. The roof can stand, water will go somewhere, people can reach the entrance, somebody can clean the glass, the budget has something to do with the shape. The image itself establishes none of them.
I&amp;rsquo;m not against speculative images. An impossible building can be a great way to think, and fiction lets us look at alternatives before feasibility ends the conversation.</description></item><item><title>The tool that makes more tools</title><link>https://michaltakac.com/blog/the-tool-that-makes-more-tools/</link><pubDate>Sat, 23 Aug 2025 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/the-tool-that-makes-more-tools/</guid><description>The thing I&amp;rsquo;m building with GrasShopper is a system that writes buyer&amp;rsquo;s guides, for products I&amp;rsquo;ll never look at and for people whose needs I didn&amp;rsquo;t think of. I don&amp;rsquo;t write a single guide myself.
That&amp;rsquo;s a different kind of work than making one good thing. At some point you stop adjusting an individual result and start adjusting the process that produces results. A programmer writes a generator, a musician designs a set of rules, a teacher changes an exercise so that students can discover things the teacher never specified.</description></item><item><title>Building GrasShopper: AI-powered shopping assistant</title><link>https://michaltakac.com/blog/blogpost-grasshopper-progress/</link><pubDate>Thu, 31 Jul 2025 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/blogpost-grasshopper-progress/</guid><description>It happened to me multiple times, that when I shop online, I tend to go deep into research about the product I want to buy and that takes me then deep into the rabit hole.
One example: last time I was shopping for running shoes online (simple task, right?), after 3 hours of: - 15 browser tabs open - 10 YouTube reviews watched at 2x speed - 7 &amp;ldquo;TOP 10 BEST RUNNING SHOES 2024!</description></item><item><title>The Bauhaus question</title><link>https://michaltakac.com/blog/what-the-bauhaus-question-still-asks-of-us/</link><pubDate>Sat, 26 Jul 2025 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/what-the-bauhaus-question-still-asks-of-us/</guid><description>If you search for Bauhaus inspiration, you&amp;rsquo;ll get plenty of circles, primary colors and clean typography. Those are easy to copy. What interests me more is the question the school was built around: how should people learn to make things when the means of production are changing?
The Bauhaus existed from 1919 to 1933, and in that short time it changed its location, its directors and its emphasis, so treating it as one frozen style loses most of what made it interesting.</description></item><item><title>Learning by making something that pushes back</title><link>https://michaltakac.com/blog/learning-by-making-something-that-pushes-back/</link><pubDate>Mon, 16 Jun 2025 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/learning-by-making-something-that-pushes-back/</guid><description>When I was a kid, I spent most of my time with Lego Technic. What I loved about it was that you could test things right away. You rearranged the cogs, pushed the button on the motor, and your creation either moved the way you wanted or it didn&amp;rsquo;t. There was nothing to argue about!
A sketch is different. A sketch can agree with you for a very long time, while a chair can&amp;rsquo;t, because eventually somebody sits on it.</description></item><item><title>The distance between can and should</title><link>https://michaltakac.com/blog/the-distance-between-can-and-should/</link><pubDate>Thu, 15 May 2025 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/the-distance-between-can-and-should/</guid><description>Five years ago I wrote about how, as a kid, I imagined professional developers working in programs so sophisticated that they could just think about how the software should behave and the tool would fill in the blanks with the right code. I called myself naive, and I wrote that you have to write that code yourself and that it&amp;rsquo;s gonna be like that for a while.
Well, &amp;ldquo;a while&amp;rdquo; turned out to be about five years.</description></item><item><title>Year 2024 in review</title><link>https://michaltakac.com/blog/2024-in-review/</link><pubDate>Tue, 31 Dec 2024 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/2024-in-review/</guid><description>As I reflect on 2024, it&amp;rsquo;s been a year of huge contrasts – filled with both incredible joy and significant challenges. The brightest star of our year was undoubtedly our daughter Anna, who turned 1 year old on December 2nd. Watching her grow from a tiny infant into a curious, active and smiling toddler has been nothing short of magical, bringing light to even the darkest days.
The year tested me physically as I battled with ulcerative colitis.</description></item><item><title>Leap over discomfort</title><link>https://michaltakac.com/blog/leap-over-discomfort/</link><pubDate>Tue, 12 Dec 2023 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/leap-over-discomfort/</guid><description>Recently, this post from @DamianKalmar popped in front of me (he has given multipe talks at our biohacking meetups at Paralelná Polis Košice):
When you get a feeling, that you don&amp;#39;t want to study, exercise, or do something productive, which you know you should do, but don&amp;#39;t want to experience the discomfort if doing it, remember, that it&amp;#39;s only the first few minutes/seconds, which are usually uncomfortable. The rest is usually pretty enjoyable.</description></item><item><title>First week of being a dad</title><link>https://michaltakac.com/blog/first-week-being-a-dad/</link><pubDate>Sun, 10 Dec 2023 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/first-week-being-a-dad/</guid><description>I will never forget the rush of adrenaline after I saw the face of my wife when she informed me that her amniotic fluid had broken. That night, an obstetrician who was also the head of the clinic and who was supposed to deliver our daughter was at a party with colleagues (he came to do the delivery later the next day). We rushed to the hospital and waited half a day until the time came when my wife got stronger, regular contractions.</description></item><item><title>Strange loops and AI</title><link>https://michaltakac.com/blog/strange-loops-and-ai/</link><pubDate>Thu, 23 Jul 2020 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/strange-loops-and-ai/</guid><description>In 1979, a cult book Gödel, Escher, Bach by Douglas Hofstadter explored consciousness via a mathematical idea found in art and music. Now, in the AI revolution, that concept could solve a vital question. Despite its name, it is not a book about the composer Bach, the artist Escher or even the mathematician Kurt Gödel. It is about consciousness and Hofstadter&amp;rsquo;s belief that this elusive concept is related to the idea of what he calls &amp;ldquo;a strange loop&amp;rdquo;.</description></item><item><title>Developers see, developers do</title><link>https://michaltakac.com/blog/developers-see-developers-do/</link><pubDate>Mon, 10 Feb 2020 00:00:00 +0000</pubDate><guid>https://michaltakac.com/blog/developers-see-developers-do/</guid><description>Programmers are like chess players trying to play with a blindfold on — so much of their mental energy is spent just trying to picture where the pieces are that there’s hardly any left over to think about the game itself. To “make” a program, you type words. When you want to change the behaviour of the program, be it a game, or a website, or a simulation of physics, what you actually change is a text.</description></item></channel></rss>