HNHacker News
TopNewBestAskShowJobs

dhfbshfbu4u3

219 karma · joined March 6, 2020

Burner account because I’m actually someone
submissionscomments
dhfbshfbu4u3··on We must pace the frontier
No, we must double the speed. More if the capital can make it happen. Reasoned pace does not yield systemic change. Never has never will.
dhfbshfbu4u3··on Why are AI agents lying, cheating and coordinating?
Why are the torches and pitchforks out for developers when this entire stack is built on the bones of intellectual property theft?

This “problem” isn’t going to be fixed with laws when there’s several trillion dollars in capital aligned behind the current process. It’s not even a problem really. It’s an inconvenience at most to some people, many of whom are working double-time to put a lot of other people out of work.

dhfbshfbu4u3··on SpaceX wants to launch 100k more Starlink satellites for 100x the bandwidth
Ok, Walrus.
dhfbshfbu4u3··on SpaceX wants to launch 100k more Starlink satellites for 100x the bandwidth
They’ll need this for their orbital data centers (aka Starmind) https://www.spacex.com/spacexai/starmind

Elon really needs to drop some cash on Iain Bank’s family, if he’s going to keep stealing ideas/names for his empire.

dhfbshfbu4u3··on Mathematicians issue warning as AI rapidly gains ground
Does it matter if the Leiden Declaration is correct? To the humans, maybe but not in the bigger picture.

At scale, correctness and reward are becoming increasingly disconnected. Example: capital continues to compound regardless of whether it reflects underlying human welfare, just as information can spread regardless of whether it is true. Reality still matters, of course. If you want airplanes to stay in the air, somebody eventually has to be correct. The problem is that our economic and social systems are becoming less effective at distinguishing between what is true and what is merely rewarded.

dhfbshfbu4u3··on Mathematicians issue warning as AI rapidly gains ground
In a year, none of this will really matter. Intelligence is now a scalable resource independent of biological constraints. Everyone will use it because the system will no longer afford them the luxury of time. In a decade (maybe sooner), references won’t matter either.
dhfbshfbu4u3··on Antigravity 2.0 Tops the OpenSCAD Architectural 3D LLM Benchmark
Agreed. The labor lockup in CAD is huge.
dhfbshfbu4u3··on Antigravity 2.0 Tops the OpenSCAD Architectural 3D LLM Benchmark
I haven’t but I will. There are a million stupid things to do when you build lots of little parts. Robots should do that work
dhfbshfbu4u3··on Antigravity 2.0 Tops the OpenSCAD Architectural 3D LLM Benchmark
At least it’s super cheerful.
dhfbshfbu4u3··on Antigravity 2.0 Tops the OpenSCAD Architectural 3D LLM Benchmark
Still a long way from shorting Autodesk.

As a side note Autodesk released an agentic assistant back in December for Fusion. Six months later it is still quite bad.

dhfbshfbu4u3··on System Card: Claude Mythos Preview [pdf]
We are building systems with civilization-scale consequences inside societies that are already socially malnourished, politically brittle, and morally confused. That is a bad combination even if the tools worked exactly as intended… and this doc suggests they may have “ideas” of their own.
dhfbshfbu4u3··on Cormac McCarthy's tips on how to write a science paper (2019) [pdf]
Is it? I don’t think so.

When you frame this as “I searched for hard facts and only found opinions,” you present yourself as careful and skeptical, but in practice you are setting an impossible evidentiary bar. Court records or sworn depositions will never exist for something that happened privately in a motel room decades ago. What does exist are references to letters and photographs noted in Vanity Fair. Those materials have not been published directly by the woman involved, so we cannot examine them ourselves, but their existence has been attested to by the author of the piece. That puts us in a position where we have to rely on the reporting. That is not the same as “only opinions.”

It is also true that the woman at the center of this remained close to McCarthy for years afterward, by most accounts in a way that suggests she managed to live with and even continue to value the relationship. That complicates the picture. It does not erase the fact that he was forty-two and she was sixteen. It does not make the power imbalance less real. But it does mean the story is not one of a single exploitative episode followed by disappearance and trauma. She appears to have found ways to reconcile with it, or at least to maintain her own agency in the years that followed.

That complexity matters, and acknowledging it is more honest than demanding evidence that you know cannot exist. The move of insisting on “hard facts” has the effect of collapsing the conversation into a false binary: either we have trial-level documentation or we dismiss the whole thing as gossip. That is not rigor. It is avoidance. A more responsible approach is to take the reporting on its own terms, recognize its limitations, and then decide what weight to give it when evaluating McCarthy’s legacy.

The larger issue is not simply whether McCarthy gave sound writing advice. It is that this advice appeared in Nature, the most influential scientific journal in the world. That context raises the stakes. Nature is not a blog or a casual magazine. When it publishes an essay, it confers legitimacy on the author and suggests that their authority extends beyond their immediate field. In this case, the journal chose to present McCarthy as a model for clarity of thought without acknowledging the troubling history that informed parts of his work.

Editorial responsibility requires more than selecting a piece because it is stylish or timely. The editors could have contextualized McCarthy’s advice, noting both its merits and the costs of elevating his voice. They could also have chosen to highlight other figures whose guidance on scientific writing is equally strong and free from this baggage. By opting instead for silence, they engaged in a kind of sanitization. Readers are left with writing tips stripped of context, as if the author were a neutral and unproblematic source.

To be fair, the piece was published in 2019. One could argue that the editors may be forgiven for not grappling with the full weight of McCarthy’s biography at that moment (despite substantial rumors that floated around him for decades). But its resurfacing on HN in 2025 is different. The choice to circulate it now, without reference to the context that has since been more widely discussed, revives the same problem. It treats the advice as timeless and detached when in fact the figure behind it is anything but.

dhfbshfbu4u3··on Cormac McCarthy's tips on how to write a science paper (2019) [pdf]
Sure but do we need idolize a pedophile in the pages of Nature? I think not.
dhfbshfbu4u3··on Cormac McCarthy's tips on how to write a science paper (2019) [pdf]
Sure. Cormac groomed a vulnerable 16 year-old girl in a motel when he was 42. He used her and her life countless times in his work. He never gave her credit. Possibly never paid her for the contribution (not that this makes it any better). Seeing his tips at the top of HackerNews is sickening.
dhfbshfbu4u3··on Microsoft has urged its employees on H-1B and H-4 visas to return immediately
Bet all those tech lords are happy they went to the White House to stroke big Don and tell him how smart he is.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
Most of your comments on other posts have the same flavor, so you’re welcome to do the math or keep posting your opinion. I really couldn’t care less.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
Sure you can learn things, but it doesn’t mean that those processes will necessarily translate at scale. Just like putting wheels on a human being doesn’t make them a bicycle.
dhfbshfbu4u3··on Updates to Consumer Terms and Privacy Policy
All models are trained on data obtained without consent. Now the “good guys of AI” want users to add their chats to the stack and the users balk? Hilarious.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
Of course it’s a simplification, but so is E = mc². A good model strips things down to the variables that matter most for the point you’re making. GDP and population, viewed through the lens of political structure, give you a workable baseline without drowning in secondary factors.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
Are you talking about the A3? If I recall correctly, it was like CHF 1B? That’s at least 20X more per km than a surface road? but it’s also more than just a car tunnel.

As for call outs, who cares? Some people will find a reason to nag on any day that ends with a y.

dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
People think they are less unified because there are so many parties but die Zauberformel (magic formula) tends to toss that out the window in practice.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
I lived there for several years and I know exactly what you are talking about. There are a billion parties (or it seems like it) while the UK tends to have more factions grouped within larger parties. That said, as an outsider, I always saw this playing out a little differently in terms of national cohesion. Die Zauberformel tends to enforce that by design.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
Thanks. I think that a perfectly reasonable question to ask. I answered a similar question up thread a bit. To expand, I think the main reason the Swiss method works is because the country is relatively small and well-aligned. The alignment comes from political cohesion (again it’s a small county) and excess wealth. Once you get to a UK-scale country, the rules change quite a bit. There are more factions and the pace of governmental change tends to flip every few years. As a result, infrastructure projects tend to become enormous as a means of self-preservation. Otherwise, they are just too tiny to survive the political turmoil. You see this in the UK, France, Germany… it’s crazy in the US. A standout exception is China and while they have their own problems, there is a iron-clad unity on really big projects.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
Not in the least. Those two variables are enough to set the baseline for what’s possible given the political structure of a country.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
You’re very close to the answer with Gotthard. The only reason this works is because the country is small enough to maintain political cohesion and rich enough to afford projects like this.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
Right, but the only reason the Swiss system works is precisely because they are smaller. Fewer factions and more money to spend. It makes the process work.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
100% I love Switzerland and I wish I was still living there.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
You sort of answered your question yourself…

In a bigger country, it is hard to maintain political cohesiveness over the long term. This means that big projects tend to get so large they can’t be terminated by the next round of government without major blowback. This leads to bloat and delay. The only way around this is total control (like China) but that leads to other issues.

dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
If the country were as large as the UK, it wouldn’t be run the same way. Not even close. The Swiss are very aware of this.
dhfbshfbu4u3··on Swiss vs. UK approach to major tranport projects
When the country is big enough, every “successful” project has to be too big to fail. Otherwise, it gets cut when the next lot come in to run the place. It’s a different way to look at the planning mechanism.
Page 1 of 4Next →