HNHacker News
TopNewBestAskShowJobs

kartoffelsaft

129 karma · joined October 28, 2024

My Blog: https://blog.kartoffelsaft.dev
submissionscomments
kartoffelsaft··on Is it possible to allow sideloading and keep users safe?
If there are rooms in your house someone else could lock you out of, do you own the house or do they?

If someone else could use your car without your permission, do you own the car or do they?

If someone could grow their own plants in you back yard, do you own the garden or do they?

If someone else could choose what programs run on your computer, do you own the computer or do they?

Saying "basic human right" instead of just "basic right" may be odd, but definitionally, owning a thing means having the right to say how it is used. Either you own it and have that right, or you don't own it and don't have that right. That's what owning means.

kartoffelsaft··on YouTube made AI enhancements to videos without warning or permission
I'm sure you can find people making arguments about a lack of quality from machines about textiles, woodworking, cinematography, etc., but digging holes? If you have a source of someone complaining about hole quality I'll be fascinated, but I moreso am thinking about a disconnecion here:

It looks like you see writing & editing as a menial task that we just do for it's extrinsic value, whereas these people who complain about quality see it as art we make for it's intrinsic value.

Where I think a lot of this "technophobia" actually comes from though are people who do/did this for a living and are not happy about their profession being obsolesced, and so try to justify their continued employment. And no, "there were new jobs after the cotton gin" will not comfort them, because that doesn't tell them what their next profession will be and presumes that the early industrial revolution was all peachy (it wasn't).

kartoffelsaft··on LunarEngine: An open source, Roblox-compatible game engine
Curious what makes you say it'd be less legally dubious than an emulator? To me, it seems this would be at the same legality as the NES emulator because you're basically 'emulating' the environment Roblox game code runs in. To be fair if that intuition's correct it'd still be legal like emulators are if they're careful.

(also not a lawyer)

kartoffelsaft··on April Fools 2014: The *Real* Test Driven Development (2014)
I don't think what we have now fits that definition. LLMs are still narrowly good at language generation, and the "many" things it's good at are things that have canonical textual / linguistic representations (code, chess notation, etc.). Much of existing AI that appears more general is hooking up more specific models together; for example, taking the output of an LLM and piping it into a TTS . Since these pieces are easily replaceable I struggle to call it one AI that can do many tasks.

Consider that LLM->TTS example's human equivalent: when you're talking, you naturally emphasize certain words, and part of that is knowing not just what you want to say but why you want to say it. If you had a machine learning model where the speech module had insight into why the language model picked the words it has, and also vision so it knows who it's talking to to pick the right tone, and also the motor system had access to that too for gesturing, etc. then at that point you'd have a single AI that was indeed generally solving a large variety of tasks. We have a little bit of that for some domains but as it stands most of what we have are lots of specific models that we've got talking to each other and falling a little short of human level when the interface between them is incomplete.

kartoffelsaft··on Why is GitHub UI getting slower?
Reminds me of a discussion that Casey Muratori and Robert Martin had over clean code and it's impact on performance... but not because of the subject matter. They were using GitHub as a medium for their discussion, and they ran into some serious lag when typing paragraphs into just the hundreds of characters (ctrl+f emoji, ~1/4 of the way through):

https://github.com/unclebob/cmuratori-discussion/blob/main/c...

kartoffelsaft··on Dear valued user, You have reached the error page for the error page
It had been written by (at the latest) September 2008:

https://googlesystem.blogspot.com/2008/09/best-gmail-error-m...

kartoffelsaft··on 'Gwada negative': French scientists find new blood type in woman
I am well below that age and I don't know my MBTI letters.
kartoffelsaft··on LLMs are mirrors of operator skill
The way it is written is to say that this is an important interview question for any software engineering position, and I'm guessing you agree by the way you say it's critical.

But by the same logic, should we be asking for the same knowledge of the language server protocol and algorithms like treesitter? They're integral right now in the same way these new tools are expected to become (and have become for many).

As I see it, knowing the internals of these tools might be the thing that makes the hire, but not something you'd screen every candidate with who comes through the door. It's worth asking, but not "critical." Usage of these tools? sure. But knowing how they're implemented is simply a single indicator to tell if the developer is curious and willing to learn about their tools - an indicator which you need many of to get an accurate assessment.

kartoffelsaft··on Why Use Structured Errors in Rust Applications?
It "must" be done in the sense that `dyn Error`, anyhow, and (if we include zig's equivalent) inferred error sets mean something different than setting up an error enum. Using an error enum makes the failures explicit, which (most) other options choose to discard because they're designed to handle any error.

Your options effectively become either writing that enum, or using a macro that reduces to that enum. Unless, of course, you're willing to have the function signature say "eh, this function just breaks sometimes. Could be anything really."

kartoffelsaft··on After months of coding with LLMs, I'm going back to using my brain
Honestly, with a lot of HN debating the merits of LLMs for generating code, I wish it were an unwritten rule that everyone states the stack they're using with it. It seems that the people who rave about it creating a whole product line in a weekend are asking it to write them a web iterface using [popular js framework] that connects to [ubiquitous database], and their app is a step or two away from being CRUD. Meanwhile, the people who say it's done nothing for them are writing against [proprietary in-house library from 2005].

The worst is the middleground of stacks that are popular enough to be known but not enough for an LLM to know them. I say worst because in these cases the facade that the LLM understands how to create your product will fall before you the software's lifecycle ends (at least, if you're vibe-coding).

For what it's worth, I've mostly been a hobbyist but I'm getting close to graduating with a CS degree. I've avoided using LLMs for classwork because I don't want to rob myself of an education, but I've occasionally used them for personal, weird projects (or tried to at least). I always give up with it because I tend to like trying out niche languages that the LLM will just start to assume work like python (ex: most LLMs struggle with zig in my experience).

kartoffelsaft··on Why Bell Labs Worked
This seems written to sound like a profound piece of wisdom, but I find it difficult to not interperet it as a very flowery way of saying "git gud." If that is indeed what you mean then that is fine, but it is still worth acknowledging that the greater competition for funding of today means scientists of today are not playing the same game as scientists of Bell Labs's time.
kartoffelsaft··on I asked police to send me their public surveillance footage of my car
The argument is not that this theoretical 3rd party would be a for-profit company, but that there's already existing for-profit companies that exist and could serve that purpose, and that the new 3rd party wouldn't see much use because of that.

They almost certainly are willing to buy hordes of data off of google/facebook/etc.; it's useful data that they're already negotiating to get. Why, then, would they be want to put in the effort for a warrant on cctv footage when the suspect's google search and maps history they're alreaty getting contain the same info and more? At best, your creating a small amount of competition for data brokers, and I question if it's that much.

kartoffelsaft··on Programmers are modern-day computers
Sure, people make mistakes all the time. But would you prefer those mistakes be sprinkled randomly throughout your data crunching, or be systematic errors?

The point that that post is making is that a machine isn't going to make a mistake in adding two numbers. It reduces arithmetic errors to 0 (unless you count overflow which can be detected), and if it didn't it would only be useful in the rare case you don't care about accuracy.

AI in it's current state does not do for logical accuracy what computers did for arithmetic accuracy; You still need to verify every output from an LLM, which I doubt you've done for the many billions of arithmetic operations that happened this second on the computer you're on right now.

edit: fixed typo

kartoffelsaft··on Add "fucking" to your Google searches to neutralize AI summaries
It really feels as though Amazon's greatest fear is the idea that you might search for something and get no results. If we show you BS you explicitly said you aren't looking for then we're giving up the opportunity of tricking you into buying it anyways.

A bit ago I was searching for toothpaste that doesn't have mint in it. This is already a pain at a brick retailer, but I figured Amazon's huge product variety would help. Turns out their search is actively malicious to negative terms because otherwise I could buy just the one thing and be done with my shopping.

I should probably set up a similar homebrew search to get around this. Purchase history is far less important to me because I don't buy much from Amazon.

kartoffelsaft··on C++ template macroprogramming versus Lisp macros
I believe a restriction for `constexpr`/`consteval` is that you can do memory allocations but they must all be freed at the end of the evaluation. Arbitrary precision means arbitrary memory usage means allocation that's the runtime's responsibility. Maybe there's a trick that gets around that (define a type at compile time that's just big enough for this specific number and return that?), but I don't see it as possible without something weird.
kartoffelsaft··on Half-Life 2: 20th Anniversary Update
For whatever reason, Ravenholm didn't particularly unsettle me. I don't know what it is, maybe expectation from the praise it receives left me underwhelmed, or maybe it's the fact that I was quite literally a toddler when the game came out and only first played it in 2019?

To be fair, I did enjoy it and could feel the intent of it being the horror section. I could see it usettling me if I played it at the time, but I see a lot of people saying it holds up really well so I still feel I've missed out on something.

kartoffelsaft··on IMG_0416
There are a few things that come to mind that make this different:

1. This isn't really privacy breaching. For someone who taps the "share to youtube" button without knowing what it means, sure, but even that is pretty explicit that you're sharing it. Not sure why the article itself says people didn't know what the button would do before tapping it, so I'd like some further explanation of this point.

2. It's opt in, not opt out. Spending time with most "normal" people has shown me that very few people give a crap about going into settings menus to configure exactly how ther data is used or collected, or otherwise switching to a service that gives them that control. When HN complains about privacy being dead, they are complaining about this apathy end how it gets exploited. This feature does not exploit that apathy.

3. This gives us something that we actually want. When most services invade your privacy, it's usually for things like advertising, targeting content, and data brokering. Things that I know I personally have a lot of issues with, and I feel I'm not alone. This button doesn't do those things, it just gives us interesting videos. So much so that most of the fascination with these videos is that you can feel the absence of those issues.

kartoffelsaft··on Trump wins presidency for second time
Trump has stated that his biggest regret from his term is that the people he appointed to various positions, while quite competent and/or experienced, would push back on ideas or plans he proposed. In other words, they weren't loyal.

The difference between this term and his previous is going to be a much stronger focus on making any position he can appoint be one that doesn't tell him no. And it looks like many of the positions he can't (the senate and likely the house) are going that way too. That, to me, makes him represent a meaningfully larger threat to the balance of power in the US than his previous term.

kartoffelsaft··on Care Doesn't Scale
Honestly sounds like that's Conway's law, but in reverse. Instead of a project's setup mirroring it's organization, the organization mirrors what they see as the ideal project setup. That's something that I wish more teams were willing to recognize as the best approach; Conway's law isn't necessarily something to be worked around, it's a tool.
← PreviousPage 2 of 2