HNHacker News
TopNewBestAskShowJobs

qrios

227 karma · joined January 19, 2013

It is not a question if, but ‚when‘ we will see what we can call AGI. The ‚when‘ can be answered if we know the ‚how‘. The ‚how‘ we know if the ‚when‘ is answered by ‚today‘.
submissionscomments
qrios··on Show HN: OctoFlow – A GPU-native programming language
Works on my computer: RTX 3090, CUDA 12.6

Interesting project! I haven't really worked with Vulkan myself yet. Hence my question: how is the code compiled and then loaded into the cores?

Or is the entire code always compiled in the REPL and then uploaded, with only the existing data addresses being updated?

qrios··on Show HN: MOL – A programming language where pipelines trace themselves
Very interesting! I'll definitely give it a try. However, the documentation link[1] isn't working at the moment (404).

[1] https://crux-ecosystem.github.io/MOL/

qrios··on The exorbitant privilege of the US brain gain is fading
https://archive.is/0PLa6
qrios··on 74% of European firms would fail without access to U.S. technology
Via certificate invalidation, or turning off services for certificate pinning?
qrios··on Mark Carney's Full Speech at the World Economic Forum
Text:

https://www.weforum.org/stories/2026/01/davos-2026-special-a...

qrios··on New .gov Portal: Foreign Funding in US Higher Education
Same for me with ISP from Austria.
qrios··on Will Creative Work Survive A.I.?
https://archive.ph/VZpQ0
qrios··on Why AGI Will Not Happen
> … what human intelligence is because intelligence is best described as a side-effect of consciousness …

Is "human intelligence" and "intelligence" equal?

And: How to become conscious before being intelligent?

Or: If intelligence is a side-effect, how often this side-effect can't be observed?

Xor: What if an intelligent being denies being conscious?

qrios··on LLM from scratch, part 28 – training a base model from scratch on an RTX 3090
"low hanging" is relative. At least from my perspective. A significant part of my work involves cleaning up structured and unstructured data.

An example: More than ten years ago a friend of mine was fascinated by the german edition of the book "A Cultural History of Physics" by Károly Simonyi. He scanned the book (600+ pages) and created a PDF (nearly) same layout.

Against my advice he used Adobe tools for it instead of creating an epub or something like DocBook.

The PDF looks great, but the text inside is impossible to use as training data for a small LLM. The lines from the two columns are mixed and a lot of spaces are randomly placed (makes it particularly difficult because mathematical formulas often appear in the text itself).

After many attempts (with RegEx and LLMs), I gave up and rendered each page and had a large LLM extract the text.

qrios··on FLUX.2
duplicate: https://news.ycombinator.com/item?id=46046916 9 days ago 117 comments
qrios··on AI Boom Is Turning Malaysia's Palm Oil Estates into Data Centers
https://archive.is/Ya9Am
qrios··on XSLT RIP
> This is actually a clever way to distinguish if the browser supports XSLT or not. Actual content is XHTML in https://xslt.rip/index.xsl

I agree it is a clever way. But it also shows exactly how hard it is to use XML and XSLT in a "proper way": Formal everything is fine to do it in this way (except the server is sending 'content-type: application/xml' for the /index.xsl, it should be 'application/xslt+xml').

Almost all implementations in XML and XSLT that I have seen in my career showed a nearly complete lack of understanding of how they were intended to be used and how they should work together. Starting with completely pointless key/value XMLs (I'm looking at you, Apple and Nokia), through call-template orgies (IBM), to ‘yet-another-element-open/-close’ implementations (almost every in-house application development in PHP, JAVA or .NET).

I started using XSLT before the first specification had been published. Initially, I only used it in the browser. Years later, I was able to use XSLT to create XSDs and modify them at runtime.

qrios··on Offline Math: Converting LaTeX to SVG with MathJax
Thnx for sharing!

With Safari (standard and tech preview) the rendering looks strange (at least). The root sign does not have a strait line at the top (for many fonts) and at least the partial derivative is not rendered as italic (for all fonts).

qrios··on Context engineering
I agree with you one hundred percent.

But: Interestingly, the behavior of LLMs in different contexts is also the subject of scientific research.

qrios··on Tell HN: Azure outage
Is that you? The same guy with the comment "hahahhahaha"[1] on "Women dating safety app 'Tea' breached, users' IDs posted to 4chan"[2]?

[1] https://news.ycombinator.com/item?id=44689366

[2] https://news.ycombinator.com/item?id=44684373

qrios··on The case for the return of fine-tuning
> Finally, companies may have reached the ceiling of what can be achieved with prompting alone. Some want models that know their vocabulary, their tone, their taxonomy, and their compliance rules.

Together with speed and const, this is from my point of view this is the only "case" for the return of fine-tuning here. And this can be managed by context management.

With growing context sizes, first RAG replaced fine-tuning and later even RAG was replaced by just a good-enough prompt preparation for more and more usage pattern.

Sure, speed and costs are important drivers. But like with FPGAs vs. CPUs or GPUs, the development costs and delivery time for high-performance solutions, eliminate the benefit most the time.

qrios··on MAML – A new configuration language
You are right. And the initial question is also meaningless as there is no difference between these two:

  {"project": "MAML"}

  {project: "MAML"}
A parser is not able to decide if '"project"' equals 'project'.
qrios··on MAML – A new configuration language
Hmm, that's just "another link from someone who uses HN as a bookmark service". User started this self-service at the end of last year. With now 60+ submissions in the last two weeks. We can expect more interesting topics.
qrios··on MAML – A new configuration language
Also not abstract. To be ‘abstract’, a language must combine several objects ‘under one umbrella’. Depicting different things in the same notation is the opposite of abstract. YACCANL (Yet Another Cargo-Cult Abstraction Non-Markup Language).
qrios··on Wireguard FPGA
> (although the license seems proprietary?)

Hm, "BSD 3-Clause License" is seems really proprietary to you?

But you are right: do the personal license in many(most?) Verilog files[1] overrules the LICENSE file[2] of a repo?

[1] https://github.com/chili-chips-ba/wireguard-fpga/blob/main/1...

[2] https://github.com/chili-chips-ba/wireguard-fpga/blob/main/L...

qrios··on N8n raises $180M
Databricks offers an engineering, management, ETL (focussed on data) platform as SaaS/PaaS. As far as I can see, this is what the revenue comes from.

n8n offers an orchestration, integration, ETL platform as SaaS/PaaS (plus self-hosted).

For many medium sized and enterprise companies such a service is as important as the data platform from Databricks. From my personal experience in a small EU-wide enterprise in finance sector, the budget for the data platform is in the lower 7digits/year. For automation and workflows multiple solutions in place. Some are outdated for years (on mainframes) or with an EoL in the next years (due to retirements in-house or at the datacenter provider).

Since decades such companies externalize their own business logic – formally running on homegrown software – to SaaS vendors. Established workflows are quite often not able to integrate these kind of "modern" integration. REST? GraphQL? WebSocket? We speak SOAP and FTP!

So, from my point of view n8n is a valid solution in a growing market.

qrios··on N8n raises $180M
> They are SaaS and not consumer products. They have no side gig like amazon or google - they have a single product in a tight market.

Like Databricks with no consumer products, no side gig, single product, but 100 billion valuation[1]?

[1] https://www.forbes.com/sites/sasirekhasubramanian/2025/10/07...

qrios··on Citizen protest halts Chat Control
https://news.ycombinator.com/item?id=45506143
qrios··on Why Microsoft Has Lower Borrowing Costs Than the U.S.
Article is now open.
qrios··on HSBC Claims Quantum Trading Breakthrough
https://archive.is/2025.09.25-054751/https://www.ft.com/cont...
qrios··on Ultrasonic Chef's Knife
I’m a former chef too and have knife skills but, no. Overripe tomatoes, grapes, carrots, meat near cartilage? No problem at all with the right tool.
qrios··on Cyberattack disrupts operations at European airports incl. Heathrow, Brussels
Some more information here:

https://cyberplace.social/@GossiTheDog/115235882835085934

qrios··on Turgot Map of Paris
It probably has more to do with different administrative areas. Cities used to have different rights. Cities could just not simply expand to external land. The reason was quite simple: the land belonged to someone else. Meanwhile, the city was independent, even if it was the capital of a kingdom (such as Paris, for example).

In Vienna, for example, the city ended behind the belt. As a citizen, you could travel back and forth between the surrounding area and the city, but different laws applied (taxes, marriage, property).

The Viennese enjoyed traveling to the surrounding countryside for leisure (winegrowers had to pay significantly less tax for serving their own products than innkeepers in the city), but the citizens did not want to live there, or there were strict regulations on moving in.

qrios··on Human writers have always used the em dash
I was wondering about this since a while. It looks weird to me as in German between the word and the em dash a space is mandatory. (At least some decades ago.)
qrios··on What Happened to Mexico City's Food Scene? Americans
https://archive.is/15SVE
← PreviousPage 2 of 5Next →