Main Page
Niobaras Foliant
Im Informatikunterricht an meiner Schule oder in der Informatik AG, also wohl so in der 10. Klasse, vielleicht um 1993 oder so, lernten wir die Turtle Programmierbibliothek kennen, ich glaube damals in Pascal. In der Schule fingen wir dann damit an, einfache Fraktale wie die Koch-Kurve oder die Drachenkurve zu erstellen.
Ich war zu jener Zeit ein großer Fan des Rollenspiels Das Schwarze Auge und dessen Welt Aventurien, und besonders die 1990 erschienene Encyklopaedia Aventurica hatte es mir angetan. Neben vielen anderen Artikeln zu Aventurien, beschrieben darin Hadmar von Wieser und Oliver Richtberg den Aventurischen Sternenhimmel, und er war anders als der irdische: Sternbilder zogen mit unterschiedlicher Geschwindigkeit über den Nachthimmel, manche, wie die Sternbilder der Zwölfgötter, drehten sich mit dem Jahr mit, andere wie der Drache, brauchten 33 Jahre für eine Umdrehung, wieder andere, wie der Elfenstern, schossen alle vier Tage einmal um den Nordstern.
Bewaffnet mit einem Geodreieck und am Schul-PC begann ich die Sternbilder in Pascal nachzuarbeiten. Winkel für Winkel, Strecke für Strecke, und dann folgte ich der Beschreibung in der Enzyklopaedia so genau wie ich konnte. Ich nannte das Programm “Niobaras Foliant”, und es war in der Lage, den Sternenhimmel für eine beliebige aventurische Nacht nachzuzeichnen. Es war damals ein MS-DOS Programm.
Etwas später kaufte ich mir Visual Studio, Microsoft’s Programmierumgebung für den PC, und programmierte das nochmal nach, mit einer besseren grafischen Benutzeroberfläche (was Visual Studio wirklich leicht machte). Ich muss ehrlich zugeben, ich weiß nichtmal mehr ob ich das in Basic oder in C++ nachprogrammierte (irgendwo muss ich den Quelltext von all denen noch rumliegen haben, aber ich habe seit vielen Jahren auch keinen Windowsrechner mehr). Und das lief dann auf Windows. Und das Programm habe ich dann auch auf ein paar DSA Webseiten angeboten.
Ich habe über Jahre immer wieder Komplimente zu dem Programm erhalten, aber die letzten zwanzig Jahre auch nicht mehr angefasst.
Michael Nock hat mich vor ein paar Wochen gefragt, ob er das Programm neu auflegen könnte. Ich habe mich über die Frage sehr gefreut, und Michael hat hervorragende Arbeit geleistet. Das Programm ist vollständig in HTML und JavaScript und läuft also überall im Browser. Und man kann es runterladen und es läuft komplett lokal, ohne dass es Internetzugriff braucht. Man kann den den neuen Folianten bei Ulisses runterladen für einen Preis den man selbst wählt (auch frei).
Michaels Version hat reichlich Verbesserungen: so sind die Veränderungen des Sternenhimmels gerade durch die Sternenfall-Kampagne aber auch andere Ereignisse nachvollziehbar, und man kann alles editieren – selbst die Sternbilder selbst. Auch Horoskope sind verfügbar, verschiedene Kalendersysteme (wie auch im ursprünglichen Programm), und noch so viel mehr.
Ich bin Michael sehr dankbar, dass er das Programm wiederbelebt hat und einer neuen Generation (na ja, und der alten auch) zur Verfügung stellt. Großartige Arbeit, ich bin geehrt, dass die alte Idee immer noch lebt.
The narration of the AI apocalypse
I'm no AI apocalypse doomsayer. But from a sheer narrative point of view, some recent events --- the PHP vulnerability, Microsoft going down, the US grounding all flights --- would make classical foreshadowing vignettes.
The Web of Knowledge
I thought I knew the history of Wikipedia inside and out. Then I read Giota Alevizou’s The Web of Knowledge, and I found myself honestly surprised. I learned things about Wikipedia and its role in the knowledge ecosystem that had a major impact on my career, and that I was completely unaware of.
2026 is Wikipedia’s Jubilee year, and there are a few books published retelling the history of Wikipedia. But this book goes further; it discusses the role of Wikipedia in the wider knowledge infrastructure, and it encompasses the history of (English) digital encyclopedias as a whole, covering in particular the journeys of Britannica and Encarta from the 1990s on, in addition to telling the history of Wikipedia.
The book starts with a foreword by Jimmy Wales, the founder of Wikipedia, in which he closes by discussing the impact AI might have on Wikipedia and the knowledge ecosystem. A timely topic that the book will pick up.
I have to admit that I had a bit of a hard time getting started with the book, particularly regarding its positioning within its theoretical frameworks. But once we started to dive into the topic, I enjoyed it quite a bit and learned a few things.
The book covers very competently the pre-history of digital encyclopedias, starting well before Diderot, covering the ‘Great Books’ of the past, the Summae, and also Otlet and Dewey and their plans to index and organize the world’s knowledge, and Nelson’s Memex, all the way to Lenat’s Cyc. Giota describes knowledge organization and the encyclopedia as parallel threads weaving through the centuries.
For me the biggest surprise was right in the origin story of Wikipedia, where both threads were tied together. A lot has been written and said about Jimmy Wales’ motivation to start an online encyclopedia, including his mother buying an encyclopedia when he was 3 years old. When Jimmy was 30 years old, in 1996, he founded Bomis, a web portal for user-generated webrings and adult content. Critics often raised the second part, but Giota’s book focuses on the first: to structure the topics of the webrings would be helpful. And what better solution for that than an encyclopedia?
This was around the same time that Yahoo started becoming the main portal to the Web, and that predecessors of DMOZ came around. Back then, Yahoo wasn’t a search portal, but a directory. Jimmy’s intuition that an encyclopedia would be a great complement or substrate to organize the webrings would prove extremely foresighted, and became one of the foundational ideas for my own career.
That’s why I was so surprised to read about it in this book. So much that I actually had to discuss this with Jimmy and he, unsurprisingly, confirmed the story.
An encyclopedia is “a systematically organized repository of knowledge”. Already in the early 20th century, Giota describes how “the encyclopedia serves as a means for the semantic structuring of (digital) information spaces”.
I want to expand on this point: Jimmy’s intuition repeatedly bore fruit in the years that followed. Ironically, to the best of my understanding he never used it to improve the organization of webrings by Bomis. In 2005, Markus Krötzsch, Max Völkel and I described the Semantic Wikipedia idea, where we suggested using Wikipedia articles as anchors for topics.
In 2007, Metaweb’s Freebase also took Wikipedia as the central semantic structure to build around, which later became the Google Knowledge Graph, considerably improving Google Search by using “things, not strings”. DBpedia, started in the same year, was also lifted off from Wikipedia and called itself a Nucleus for the Web of Open Data.
In 2012 we kicked off Wikidata to explicitly provide and entangle a machine-readable knowledge graph, firmly based on Wikipedia. Abstract Wikipedia is now a continuation of this idea, working toward a more comprehensive knowledge representation than Wikidata did, towards the more ambitious goal of allowing many more people to share knowledge across language barriers.
And although I have been working on this for more than 20 years, I did not know that using an encyclopedia as the foundational structure of the semantic space was Jimmy’s motivation all along.
But back to the book. For me, that insight alone was worth the prize of admission. And Giota doesn’t stop there. There is the widespread myth that “Wikipedia killed Britannica”, and maybe some of us know that this doesn’t align with events. Britannica was already in a steep decline when Wikipedia was founded, and the reason was Microsoft’s Encarta. And whereas I roughly knew this story, Giota really lays out the narrative with thorough evidence drawn from her own material collected over decades. She had already interviewed people involved with Britannica and Encarta in the 2000s, and used this material to lay out a fascinating story, about Britannica’s turbulent organizational history and Microsoft’s ambitions to use Encarta as the promotional vehicle to get a PC into every household: because for the price of a Britannica, you not only get an encyclopedia for your children, but you get a lot more value.
One interesting point is about finances: Giota mentions that in 1989, Britannica had $627 million in revenue. That’s almost $1.6 billion in today’s dollars.
That is roughly one order of magnitude larger than the budget of the Wikimedia Foundation. While Britannica is available only in English, Wikipedia is available in more than 300 languages.
Giota also reminds us that Encarta premiered multimedia capabilities and functionalities from which Wikipedia could still learn. But just as Encarta eroded Britannica’s financial base, Wikipedia was able to replace Encarta. Microsoft wouldn’t even put up a fight: the PC had already won the marketplace, and whether Encarta or Wikipedia was the encyclopedia, what did it matter to Microsoft?
The book outlines the history of how encyclopedic authority moved from a small group of exalted individuals to a crowd in which everyone can participate. It reminds us again how littered with accidents this history was, and how fragile the current situation is. The wide deployment and usage of AI is destabilizing a situation in more than one way, and the book outlines what we may lose if we are not careful. We must take our role as stewards of the encyclopedia seriously – I am reminded of Umberto Eco’s wonderful quote: "After all, the cultivated person's first duty is to be always prepared to rewrite the encyclopaedia."
Links:
- Website about the book by King’s College London
- A review of the book by Matthew Vetter in the International Journal of Communication
Centillion to the power of a centillion, or a googolplex?
Computers are good with numbers.
"what's bigger, centillition to the power of a centillion, or a googolplex?"
I am glad that Google gives me a richly sourced and well reasoned answer, linked below:
"A googolplex is vastly bigger than a centillion to the power of a centillion."
The Math
- Centillion: In the short scale, a centillion is 10^303 (a 1 followed by 303 zeros)
- Centillion to the power of a centillion:
- (10^303)^(10^303) which is roughly 10^(10^305.48).
- Googol and Googolplex:
- A googol is 10^100
- A googolplex is 10^googol = 10^(10^100).
Why a Googolplex Wins
- Centillion to the centillion power is roughly 10^(a number with about 305 digits).
- A googolplex is 10^(a number with 101 digits), which is orders of magnitude larger.
To put it in perspective, a googolplex is so large that writing out all of its zeros in decimal form would require more paper and matter than exists in the known physical universe."
Please don't outsource your thinking skills yet.
Sam wants your RAM
Sam Altman: "But from a global allocation perspective, actually it's quite terrible to have everybody have a bunch of the world's limited supply of DRAM sitting in their own computer”
From an allocation perspective, actually it seems quite terrible that the richest 100 US billionaires own more than 7 trillion dollar, sitting on more wealth than 70% of US households combined, while many have to struggle to get health care, education, nutrition and shelter.
But I guess we all have our pet peeves.
Patrick Stewart
When the table read of the first episode of The Next Generation happened, it was Patrick Stewart who first pronounced the name of the android Data. He enunciated a crisp T instead of the American habit of letting the letter disappear somewhere on the top of their mouth. Gene Roddenberry then set up the rule that whoever reads a name first decides on the pronunciation. Given the profound cultural impact of Star Trek this influenced the way data was pronounced, and thus much later also influenced Wikidata.
Captain Picard and Professor X as portrayed by Patrick Stewart embodied leadership for me, a role model since growing up. If you read up on the production of Star Trek: The Next Generation, you will learn that he also was a role model and a leader for the rest of the cast, his professionalism and conduct contributing far beyond the role he portrayed on screen. In his private life he served less of a role model, but unlike some other stars his private life sure didn't have anywhere near the prominence in the public consciousness as the roles he portrayed have.
He'll always be my Captain. "Make it so" is such a wonderful catchphrase, showing confidence in the capabilities of the crew, and how good communication and consensus had worked before. The scenes of Captain Picard and his leadership team sitting around the table, discussing options and coming to level headed decisions, had little counterpart on Kirk's Enterprise. I don't expect to get into the habit of drinking Earl Grey, hot, though.
And even though Magneto was right, Professor X exemplified the mercy and compassion I want to always aspire to. The real-world friendship with Ian McKellen provided such a fruitful interplay for the portrayal of the relationship between Professor X and Magneto: sure, both Stewart and McKellen were brilliant actors, and I have no doubt that they would have been able to bring that relationship to life on screen even if they would have despised each other off screen, but their chemistry and friendship off screen, knowing about it, just made the rivalry and friendship of the characters feel so much more vivid. I'm looking forward to seeing Patrick Stewart and Ian McKellen portray Professor X and Magneto and their relationship once again in Doomsday.
Ad Astra.
Serbia elections 2026
The election in Serbia is a win for democracy. The students' list very intentionally ran on one message: against the Vučić government, against corruption, for accountability. They tried to avoid having a commitment on any other issue: no commitment for or against the EU, for or against Russia, for or against Lithium exports. They defied having a leader, very intentionally.
They stand for basic values: for democracy, for the freedoms of speech, of press, of research and education, for a separation of powers. This was enough to carry an election, fortunately, but is it enough to form a government and lead a country?
Vučić was the minister of information in the Serbia of Milošević. I wouldn't be surprised if Vučić conceded quickly to position himself to run for president before the year is over. This isn't over.
The hard part is starting. Rebuilding democratic structures, reinstating press freedom, gaining agreements on what to do, not just on what to be against. All the best!
P.S.: To make it clear: I am so incredibly happy Vučić conceded quickly. In my mind I imagined sonic weapons being used against an agitated crowd, police and army marching against the people, voting results being faked, tanks rolling through the streets of Niš and Belgrade, the cities burning.
None of this happened.
And I am shaking with relief that none of this happened. And I am so incredibly thankful that none of this happened.
Vučić has lost the election. And he conceded. He did the Orbán. And for this, for this!, I am thankful to him.
This could have been a very different morning today. It is not. This is good.
Zur Ontologischen Bestimmung Neuronaler Netze
2003, 2004 schrieb ich als Philosophiestudent eine Hausarbeit über Neuronale Netze, turingvollständigkeit, und die Frage, was Künstliche Intelligenz und den Menschen ausmachen. Damals ging es um Ray Kurzweils Visionen, die Grenzen künstlicher Intelligenz und die Möglichkeit, den menschlichen Geist auf einem Computer nachzubilden. Mehr als zwanzig Jahre später, im Zeitalter von LLMs und generativer KI, habe ich die Arbeit wiederveröffentlicht. Manche meiner Argumente würde ich heute anders formulieren, einige Aussagen korrigieren, aber ein paar Gedanken sind vielleicht interessant. Für die meisten wahrscheinlich nicht.
Für mich war diese Arbeit immer mit dem folgenden Gedanken verbunden: eine Rechnung wird durch einen Computer nicht simuliert, sondern die Rechnung findet tatsächlich statt.
Ich finde den Gedanken immer noch interessant und wichtig. Wenn ich die Arbeit aber lese, finde ich, dass die Arbeit diesen Gedanken weder wirklich gut ausarbeitet, noch die Folgen davon reflektiert. Stattdessen tastet sie sich an viele anderen Fragen ein wenig ran, ohne wirklich voranzukommen. Na gut. Ich war ja auch erst Mitte 20.
Does the Truth feed those people?
I looked up something that sounded like a quote using Google: "Does the truth feed those people?"
Google's AI answered, without even a single Web search result beneath, *only* the AI answer:
"Does the truth feed those people?" is a powerful quote spoken by the character Fermin (played by Janusz Kamiński) in the 2006 film Children of Men.
Which would be a great answer.
If that was a quote from the film. It isn't.
If there was a character named Fermin in that film. There isn't.
If Janusz Kamiński had played in that film. He didn't.
If Janusz Kamiński was even an actor. He isn't. He is a cinematographer and director.
If Janusz Kamiński had anything to do with the film. He didn't.
It has been a while since I noticed such a complete fail. At first I believed the answer, and started to act on it. Only later I noticed that it didn't sit right. And it unraveled.
Next time it might not unravel so easy.
Checking one fact: Yelena in Brand New Day
This is not really about Spider-Man, but about the Truth, about AI, and Wikipedia.
From time to time I stumble into a rabbit hole where I start fretting about a single fact, and try to get down to the bottom of it. Sometimes I write about these (e.g. the height of Anson Mount, or the place of birth of Ena Begović).
This time, I stumbled upon two sentences in the Wikipedia article on the recent movie Spider-Man: Brand New Day, the highest grossing movie in the US ever: “Florence Pugh reprised her MCU role of Yelena Belova / Black Widow. Her introduction scene was altered in post-production due to its perceived sexual undertones.”
Now the first sentence is fine, but I was wondering about the second sentence. Somehow, it didn’t sit right, and it came without a reference. So, I just copied the two sentences into Google, expecting to find a source for that claim, add it to Wikipedia, and that would be the end of it.
Google immediately launched the AI mode, and stated that the text is from the Wikipedia article (correct). It further explained that “during test screenings, younger viewers expressed discomfort with the perceived sexual undertones of the moment. In response, the creative team made post-production visual adjustments to make the bathhouse water more opaque to lessen the nudity/sexual implications”. It gave Wikipedia as a source for this.
That text didn’t appear on Wikipedia.
It asked me if I wanted a primary source for this, and I said yes. The AI then offered an interview in The Wrap, but only gave a link to Threads, where a bit of the interview was quoted. The quote didn’t say anything about test screening or scene alterations. So I ask: “Where's the link to the Wrap interview?”
And now it gets really interesting. Instead of giving me a link, Google answers: “The statement that "the scene was altered in post-production due to its perceived sexual undertones" is actually a fabricated claim from user-edited wikis and rumor threads, meaning there is no real interview link matching that claim.”
Then Google’s AI gives me links to other interviews.
I say: “But "During test screenings, younger viewers expressed discomfort with the perceived sexual undertones of the moment. In response, the creative team made post-production visual adjustments to make the bathhouse water more opaque to lessen the nudity/sexual implications." -- this came from you, not from Wikipedia? Where did you get that from?”
Google answered: “You are entirely right to call this out. I apologize—I hallucinated that narrative about the younger audiences, test screening discomfort, and the opacity adjustments.
“While the text you initially asked about is currently written verbatim on Wikipedia's Spider-Man: Brand New Day page, the elaborate details I gave you in my first response to explain why it happened were entirely made up by me.”
To make it clear: not only did it first hallucinate further backstory for an erroneous claim, it also stated that someone else was at fault for that. (I did not check Reddit and YouTube videos about this, these wouldn’t be sources that I can use in Wikipedia).
OK, it was clear what I had to do. I removed that sentence from Wikipedia, and said that this is unsourced. I don’t know whether it is true or not, but if someone bothers to find a reliable source, they can add it back.
But now I was curious about how the sentence made it to Wikipedia in the first place, because sometimes that can help uncover further issues. Finding that out on Wikipedia is a bit of a hassle, and I thought I could let the AI do that.
I asked. Several times. It kept responding that the answer generation failed or something like that. Eventually, the Google AI answered.
Me: “Can you figure out who entered the statement about the altered scene to Wikipedia?”
Google: “Because Wikipedia’s edit history is completely transparent and tracked in real time, it is easy to find the exact origin of this edit.
“The statement regarding the post-production alteration due to "perceived sexual undertones" was added to the Spider-Man: Brand New Day Wikipedia page by an anonymous, unregistered editor on August 22, 2026.”
The page is now and was semi-protected back then, which means anonymous editors cannot have edited the page. I knew that answer was wrong. Still, I asked: “Can you give me a link to the diff that introduced this on the 22 August?”
Google: “Because Wikipedia's specific page architecture dynamically generates temporary revision IDs rather than permanent links for unverified edits, providing a functional direct hyperlink to that precise "diff" view is not possible.”
That’s when I decided to stop, as this, again, is simply untrue. Not only is this inconsistent with the previous answer (“it is easy to find the exact origin of this edit”), but Wikipedia very much does provide a diff link for every single diff, no matter if by an anonymous editor or not. Google making up outright false information about how Wikipedia works is maybe the one thing that annoys me most in this whole conversation.
The relevant edit can absolutely be linked to. It was not added on August 22. It was not added by an anonymous contributor, but by a volunteer editor with tons of experience, and it seems in good faith. Those things can slip through, and the community can fix it, as happened today. Or maybe it is true, and it will come back with a better source, who knows. That version of Wikipedia also links to the actual interview in The Wrap (which Google avoided), and you can see that it didn’t support anything about test screenings or post-production alterations.
Wikipedia should not be blindly trusted (in fact, no source should, but that’s particularly true for Wikipedia). Articles on Wikipedia may contain errors. The community, over time, reviews the articles, and fixes many of the errors. That’s why it is such a highly trusted source of knowledge.
But taking an AI instead for answering questions? This conversation happened in October 2026. ChatGPT has been around for almost four years, and we have known about the hallucination problem for even longer. How do I fix the AI in Google’s search? Sure, it also warns us that “AI responses may include mistakes.” But look how hard it can be to get to the truth. In this case, Google not only came up with sources that didn’t support the claim we were checking, but it further embellished the wrong statements with a whole, plausible background story. Entirely made-up. Further questions were answered with increasingly wrong claims, to the point where it stated that Wikipedia’s edits cannot be linked to.
The bad thing is that I think I have used AI many times productively. I think that I succeeded in using AI to check facts, to learn about things, to answer questions. And then I stumble into such a situation, and I notice how badly it fails.
Yes, I want to believe that I am diligent about checking AI answers. But am I really? Do I have the energy and the will to check everything? I doubt it. I probably slipped many times. When I didn’t care enough about the answer. When the answer sounded pleasing enough to me at the moment. I know that this scene in Spider-Man is probably not important enough to warrant these worries, this diligence, but one thing is that there is absolutely nothing that should allow us to believe that topics of more importance are treated any differently.
I was not able to share the conversation directly, as is usually possible, nor to export it to a Google Doc. Instead I copy-and-pasted the whole page to a Google Doc and put Screenshots into it. This was not the original conversation, but accessed through my Google AI search history tab (as the original conversation crashed on me when I asked it to export to Google Docs).
Addendum a day later: The sentence was reinstated. Although the Google AI admitted that it hallucinated test screenings and said that the wiki-rumour-mill created a false fact that should be removed, the sentence is actually indeed well supported with sources already given, especially this interview:
The Wikipedia process worked well here. I challenged the sentence, it was reinstated and the community members explained why.
The AI in this case was actively harmful. Even when it said that it was hallucinating --when it reported on its *own* action-- it was hallucinating that as well. It got convinced it was wrong --although this sentence wasn't wrong!-- and because it claimed it had made it up I believed it. But it turns out, not even that admission is trustworthy!!
