Ingmar Bornholz

Für Enterprise AI sind die Modelle der unwichtigste Faktor: Daten, Prozesse und Governance machen den Unterschied. bornholz.com/2026/05/das-mo…

@RyanRozbiani Anyway - I want to have the same Botox shot as the man on the right.

Admiral Robert Harward is GOING VIRAL Here are BOTH CLIPS UNEDITED of the interview from 5/18 and yesterday 5/21 People are saying it is NOT HIM it is someone else WEARING a mask. To me it looks and sounds like two different people, but both are of him Watch and tell me what do you think?

@Chris_OH58D AI will never be able to do this 😉

You can really see that this awesome sounding machine does not really want to turn. It was a handful to drive and not even that fast. Those were different days. We may heroing them now, but.....It was kinda stupid and the result was unavoidable. The car is a legend. The man is a legend. No front.

@gdb True, Model is commodity. Data and Governance are the main product.

the model alone is no longer the product

@othylmann Tagesschau gestern Abend zum Thema startupverband treffen Berlin oder so ☺️

@fredl Wo was denn jetzt? Ich kenn nur ein altes startplatz Logo Collection Bild :)

@othylmann da denkt man Nix Böses, schaut seit Jahren mal wieder Tagesschau, und das GiantSwarm Logo brüllt einem ins Gesicht 😅

AppleTV shows seem to be the best atm: friends and neighbors, fa mankind, shrinking, morning show. Great stuff.

@pmoe 🥲

@fredl hurr durr

That's German Efficiency, ya know.

So, Apple has conducting a study on how good LLM and LRM are performing in logical games. Spoiler: not really good. LLMs are "non thinkers". Background: bornholz.com/2026/05/das-mo…

@HowToAI_ yes, that's the fundamental problem, You have to put ontology based knowledge-models besides language models. Only way to get closer to something, that actually "knows".

Apple has published a paper with a devastating title: “The Illusion of Thinking” It argues that AI models, no matter how brilliant they may seem, do not understand what they are doing. They do not solve problems. They do not reason. They merely generate text word by word, trying to sound coherent. Apple tested the most advanced reasoning models in the world on controlled puzzle environments. They tore open the internal "thinking" traces. What they found shatters the narrative that we are getting closer to AGI. Current models don't scale with complexity. They have a hard mathematical cliff. And they do not degrade gracefully. They collapse. But here is the most unsettling part. When a problem gets too complex, the AI doesn't use its remaining compute to try harder. It just gives up. Its reasoning effort actually declines. It stops thinking and starts guessing. Then Apple ran the experiment that closes the casket on the reasoning debate. They gave the AI the exact, step-by-step algorithm to solve the puzzle. The cheat codes. All the AI had to do was follow the instructions. It couldn't do it. Performance didn't improve at all. When the complexity gets high enough, these models fail because they cannot actually execute a logical sequence. They are not reasoning. They are just pattern matching. When you give them a simple problem, they overthink. When you give them a hard problem, they collapse. Paper: The Illusion of Thinking, Apple, 2025

@PalmyrPar Too much Italian weddings.

Why is he turning into a Mediterranean retiree

@ClementDelangue @pmarca Models become a commodity. The moat is in the data it is able to work on. The model and hardware invest race is pointless.

I remember when people were saying "It's useless to open-source big models because nobody will be able to run them fast"....

@bored2die That’s because German cars look awful these days. Missing Italian car designers like Bruno Sacco

Its crazy that Toyota is making better looking cars than the germans now

@SebJohnsonUK @Logist_BCB The data we have is better. Because it’s unique. That’s what counts. Models become a commodity.

There is a big conversation happening right now about sovereignty in European AI. However, Europe will only win if it is good enough to stand on its own. Paris is quickly becoming one of the most exciting cities for AI in Europe, and the ecosystem behind it is growing fast. @joinstationf is at the heart of that, led by GOAT @roxannevarza. It's the world's biggest startup campus based in Paris, housing 1,000 early stage companies across 30 programmes, with startups raising an INSANE €1.5bn last year. The companies coming out of there are not just French success stories: @MistralAI, @huggingface, @amilabs and @dust_tt are competing at the highest level globally, and the talent pool that built them is now staying in Paris and building again. What's coming out of Paris right now is putting European AI on the map 🔥 This quote was taken from my interview with Roxanne, which you can find wherever you get your podcasts.

Interesting point Bezos has her: in Germany the lower half of net-income pays only about 6% of the whole income tax. That's more in the US, but still: why not making the lower half complete tax-free? You can find the 6% in efficiency. No income tax would make a huge difference for the lower incomes.

Water jets used to minimise dust spread during a demolition in Germany

Following how of Steve explains the showroom is a blueprint on how to sell "difficult to explain" products. I especially love the red phone ;)

@Parisianaes1 The reporter was Urlich Wickert de.wikipedia.org/wiki/Ulrich_Wi… love his work.

In 1984, a German reporter in Paris shared this survival tip for crossing Place de la Concorde: “Walk steadily and never look at the drivers. If you do, they’ll assume you’ve seen them and won’t stop.” Love this video 😅👌

The narrative perfectly fits the attitude, things have to be easy. That’s why one framework has been layered over the other in recent years-and in the end it’s all JavaScript. Real code in real enterprise is hard. AI can certainly support and help, but it’s crucial to understand what is actually happening.

One of the problems with AI coding is that the narrative on X (and other social media platforms) is mostly set by people who don't have the deep coding and software engineering experience of the likes of Bjarne Stroustrup. Meanwhile the fundamental problems remain unaddressed: 1- AI generates super-human volumes of code 2- The code can be buggy, have security holes, be inefficient, etc. 3- The person who owns the code can be mostly unaware that such problems exist, so they won't even go after fixing them 4- The people who can actually fix the code (i.e., the software engineers who understand design, architecture, security best practices, etc.) are so overwhelmed that some of them will give up Meanwhile, AI companies are constantly pushing the narrative that you don't need to look at the code and the AI will fix everything itself. What they don't tell you is that if your code fails, you'll be held accountable, not them.

I rejoined x a week ago, and had seen like half/half more democratic/republican postings. After just interacting a few times, I know only see democratic leaning posts. It’s a bit sad, that my filterbubble is being created that fast. Should be more even. We always need to see all sides of the table.