HNHacker News
TopNewBestAskShowJobs

nomonnai

66 karma · joined January 3, 2024

submissionscomments
nomonnai··on LLMs understand nullability
It's a prediction of what humans have frequently produced in similar situations.
nomonnai··on A university president makes a case against cowardice
It's not an equation in what it does to people. Yes, abduction is worse than being yelled at.

However, it's pointing out that the general principle has been established: "People whose opinion I don't like can be banned from society." At first, it's only removing individuals from public discourse (cancel culture), then it's removing people physically (deportation).

This is always the endgame of eroding core liberal values. This has been pointed out to the illiberal left time and time again, to no avail.

nomonnai··on Open Euro LLM: Open LLMs for Transparent AI in Europe
It's not only Poland. I've had the misfortune of suffering through a couple Horizon projects, with partners from many European countries. Same experience as you had. "These are not serious people" sums it up.
nomonnai··on Go Should Sometimes Be a No-Go
Yeah but what does simple mean? I struggle with that a lot. In my experience, keeping it simple means not being flexible when requirements change. Adding new features becomes tedious or even a mess. Keeping things simple is an art, certainly not an easy one.
nomonnai··on A wonderful coincidence or an expected connection: why π² ≈ g
As a German speaker, the claim "more than one prenominal genitive is impossible" seems interesting but perhaps not entirely accurate. As a non-linguist I probably misunderstand the meaning of prenominal genitive and might be missing your argument's point. For the overarching discussion we should note though the great variety of refering to "more than one element" in German.

> Anna's mother's sister's dog's house

> Das Haus des Hundes der Schwester der Mutter von Anna.

It's difficult to parse though. We only get to know about Anna at the end of the sentence. Consequently, we avoid such sentences or use workarounds.

> Von Annas Mutters Schwester dem Hund sein Haus.

If you ever use such a construction German speakers will correct your bad language but they will perfectly understand the sentence's meaning.

A quick search brought me to this presentation: https://www.linguistik.hu-berlin.de/de/staff/amyp/downloads/...

The take-away seems to be: "German does not work like English" (slide 8) so be careful when comparing PreGen in German and English.

nomonnai··on Xkcd: CrowdStrike
What came to my mind whithout seeing the hovertext. I'm old.
nomonnai··on The Objects of Our Life (1983)
> I have found that audiences genuinely appreciate when you speak powerfully and simply

Unfortunately, that only works if the speaker has interesting ideas to share. Often, obtuse language hides the lack of interesting ideas. Sometimes, there are no ideas at all.

Saying "most of our ideas were pretty bad" only works if an interesting idea comes afterwards. But such every interesting idea requires a lot of work; work that many speakers have not performed.

nomonnai··on Show HN: Create Music with R
In base R, `+` is the usual arithmetic operator for numeric and complex vectors. I'm not a super R expert but I'm pretty sure the addition operator is overloaded here. It simply means:

    add(Music(), Meter(4, 4), Line(...), Line(...), Articulation(">", 1))
Edit. Found the definition in the source [1]. I think the approach is similar to Go's approach to method definition [2]. Methods are defined on types, i.e., methods are functions with special receiver arguments. The infix notation instead of prefix notation is just syntactic sugar.

[1] https://github.com/flujoo/gm/blob/69d639be86b0cef80d815a73a6...

[2] https://go.dev/tour/methods/1

nomonnai··on AI system achieves 96% accuracy in determining sex from dental X-rays
Are you doing that in a professional capacity or is this more of a hobby?
nomonnai··on Google Scholar search: "certainly, here is" -chatgpt -llm
> surge in superlatives like "vital", "essential", "crucial", and "pivotal"

This is very interesting. It could also be an effect of Grammarly, which suggests these replacements for the generic "important." Perhaps it's a combination of several effects.

nomonnai··on "Help, I see a problem and no one is prioritizing it "
Serious question: How do you avoid being blamed after all? I've observed reactions like, "It looks like you are just incompetent." and, "Whatever, just find a solution."
nomonnai··on Google CEO Calls Biased AI Chatbot Responses Unacceptable
I believe these issues are instances of the frame problem [0]. Specifying the effects of an action is easy ("show more diversity"), but specifying non-effects is hard to impossible ("do not show more diverse Nazis"). Computer science and logic have worked out how to avoid side effects in formal systems, but the real world is a different animal.

[0] https://en.wikipedia.org/wiki/Frame_problem

nomonnai··on Life Aboard a Nuclear Submarine
Life on submarines differs substantially between nations and types. Nuclear boats are pretty large, operate in the ocean, and spend their deployment below the surface. In contrast, non-nuclear boats (diesel-electric, hydrogen fuel cells) are primarily small, tend to operate in littoral zones, and may spend time on the surface regularly.

In any case, WW2 submarines were torpedo boats only diving when necessary. The battery capacity of a German Type VII or an American Balao class was measured in hours, not days. Before snorkels reached operational status, boats had to completely surface for recharging; that is not the case anymore.

I think life on a Type VII was much nastier than on a modern Type 214. The engine was extremely loud, exhaust fumes were regularly in the air, and it was cold and incredibly cramped. Living conditions are much better on modern subs, even if space is still at a premium, especially on smaller boats.

What surely remains is the sense of camaraderie because of a shared fate; each sailor's life depends on the boat functioning as a unit. In that sense, nothing has changed.

nomonnai··on Earth just experienced its hottest 12 months in recorded history
True. However, as I experienced it, there was negligence regarding a base load problem among those favoring renewable energies.

I got into emotionally charged arguments for even asking, "Will Germany produce enough electrical power after shutting down nuclear power stations?" People told me base load was an outdated way of looking at the energy market and a conservative talking point to justify cutbacks in subsidies for renewables. An often repeated argument was that Germany has been a net exporter of electrical power for many years, so reducing capacity should not be a problem.

Personally, I feel that neither side engaged in an honest public debate. I remember very well a leading Green politician, Jürgen Trittin, declaring that the transition towards renewables would cost each German citizen as much as an ice cream cone. Yes, politics must create positive momentum, but being off by orders of magnitude signals fundamental incompetence. The usual counter is, "Had everything gone according to plan, it totally would have worked out." That is childish and not a way to do serious politics.

nomonnai··on I Got Cancelled [video]
The Facebook page of an incredibly talented and passionate bass player got deleted without apparent reason. I suppose it is the nth reminder not to trust big platforms. However, there are only so many ways to reach a large audience, so there seems to be little choice.

I used the video's original title; I hope it is not link bait.

nomonnai··on China Set to Lose 60 Percent of Population by Century's End
Peter Zeihan has entered the chat.

On the one hand, I found Zeihan to be hilariously wrong about details I am familiar with, such as calling all flat-top helicopter carriers "jump carriers." This video examining Zeihan's central claims regarding China found that most of them need some serious qualification (https://www.youtube.com/watch?v=XupM5_zHDbM). Here's a Reddit thread discussing many more instances of factual errors (https://old.reddit.com/r/geopolitics/comments/hv46xb/what_ar...).

On the other hand, who cares whether the USS America has a ski jump? It still can launch the F-35. Do we miss the forest for the trees with such details? Before listening to Zeihan, I had no idea about how serious China's demographic time bomb was. It help's that he's super entertaining; a gifted story teller.

Do small factual errors add up to being wrong about the big picture? Or is a disregard for details a great means to filter the signal from the noise?

Reddit's verdict: "You know those early paleontologists who took actual dinosaur bones and made dinosaur skeletons that were sometimes accurate and sometimes batshit? He's like that but with geopolitics."

nomonnai··on 'Obviously ChatGPT' – how reviewers accused me of scientific fraud
This is an alarming development. Nevertheless, as a scientist, I believe it's an effect of a misguided culture of trust. This culture has, paradoxically, allowed mistrust of the sort described in the article to fester.

> So much of science is built on trust and faith in the ethics and integrity of our colleagues.

This is where things went wrong. We need a culture of "show me," not "trust me," the core of critical rationalism: Establishing the convention that checking each other's work is the only way to advance our understanding of the world.

Figuring new things out is an error-prone process. Sometimes, these errors are not known to a researcher; sometimes, they are known but deemed non-critical; sometimes, a person has ulterior goals that would be endangered by acknowledging and correcting the error. I don't judge. We've all been there.

However, things have been swept under the rug for far too long. If large-scale attempts can only replicate 50% of the studies investigated, published results from psychology cannot and should not be trusted without further checks (https://journals.sagepub.com/doi/10.1177/2515245918810225). The problem may be less acute in other fields, but perhaps only due to a lack of scrutiny (https://journals.plos.org/ploscompbiol/article?id=10.1371/jo...).

Absent an established process to verify a publication's central claims, substitute measures are used, such as "sounds like ChatGPT to me." This is an effect, not a cause, of a culture that values "trust me" over "show me."

nomonnai··on 'Obviously ChatGPT' – how reviewers accused me of scientific fraud
I am afraid you're right. On the other hand, in my experience, writing an academic text is rarely straightforward, with a lot of forth-and-back involved. Ultimately, a human must decide whether the diffs look more like such a pattern or a script-generated one.

Edit. Btw, you can use Git with Word's docx-format out of the box. Docx is a zip folder with XML inside; git-diff will show text changes. Further, there's git-merge-driver for merging various file formats (https://github.com/Praqma/git-merge-driver).

nomonnai··on A Visit to the Museo Nacional de Antropología, Mexico City (2023)
Yes, it was atrocious, but humans are going to do human things. Cortez's conduct was not so different from how the Aztecs had conquered the region in the preceding 150 years. In any case, germs from Europe, Asia, and Africa would have wreaked havoc no matter what. The Americas would have had to be isolated until the discovery of vaccines.

Still, hard to imagine how much art, music, philosophy, world view in general was lost and where it would be today if American societies could have developed into nation-states.

nomonnai··on How to get rich in the 21st century
There are numerous counterexamples to the factors you mention.

Czechia (Czech Republic) is a strongly atheist country and doing quite well. None of the East Asian countries are Protestant or Jewish. Italy's GDP, a Catholic country, is on par with Canada's; Poland, another Catholic country, is enjoying robust growth; Ireland, also Catholic, is doing fantastic.

South Korea's growth was initiated under a military dictatorship, as was Spain's. The country with the highest GDP per capita, Monaco, is hardly democratic; Singapore was ruled by a dictator for a long time; Vietnam, a socialist republic, is on a remarkable upward trajectory.

nomonnai··on How to get rich in the 21st century
The article is about how countries get rich, i.e., boost their GDP. Of course, people in these countries will benefit from a growing GDP, but it's a different subject matter than amassing riches as an individual.

China's (12.7) and India's (2.4) GDP per capita are very low compared with South Korea's (32.4) or Japan's (33.8). The average Chinese or Indian citizen is relatively poor compared with their peers in developed countries. Changing this requires long-term sustained growth, which is difficult to get right.

China is greatly challenged by its demography and super-power ambitions with stagnating growth. India's problems have much to do with bureaucracy and lack of infrastructure, but it has shown robust growth in recent years. However, as difficult as India's problems are, China's demographic outlook and its political implications are nightmarish.

nomonnai··on Coal and fossil fuel use in Germany at a record low for electricity in 2023
Fellow German here. I share your sentiment 100%. I quit consuming standard news early last year and replaced it with reading primary sources if the topic was substantial. Further, there are capable experts on Youtube. However, cutting through the noise and hyperbole takes time, and sometimes experts do not realize having left their field of expertise.

By now, I fully agree with Neil Postman's verdict that news primarily exists to entertain. Many discussions could be decided by looking at a chart if the discussion's participants had the will and competency to do so, which, I am afraid, is regularly not the case.

Regarding the production and consumption of electrical power, I find www.smard.de/home/marktdaten a great tool. Before using it, I had no idea of basics such as the order of magnitude of power production and consumption or the meaning of units like Watt hour. In effect, quitting news helped me to expand my horizons far more than I expected.

nomonnai··on Coal and fossil fuel use in Germany at a record low for electricity in 2023
No, Germany has been a net importer in 2023. Whereas in 2022, 27.5 TWh were exported, in 2023, 8.6 TWh had to be imported. The summer was particularly disastrous, with 5.5 TWh imported in August alone. Germany consumed 483 TWh in 2022 and 457 TWh in 2023.

Image of net export (2023) by month: https://ibb.co/jDkg9sF.

You can query and visualize these data on the official site of the Bundesnetzagentur (Federal Network Agency): https://www.smard.de/home/marktdaten. It's a capable tool, you should check it out. Once using it I realized that news reports cannot replace looking at the time series data.