HNHacker News
TopNewBestAskShowJobs

karaterobot

21,592 karma · joined October 22, 2009

         __
    (___()'`;
    /,    /`
    \\"--\\
submissionscomments
karaterobot··on DuckDuckGo now lets you hide AI-generated images in search results
There's a group of people for whom AI images seem to just be a trigger, and they simply never want to see them. So having a filter to remove those images makes sense. I don't really have that issue with AI images, but I often do want to have some way of filtering photographs from illustrations—be they drawings by humans, or images generated by AI. And that could go either way: maybe I want to see what an actual baby peacock looks like in nature sometimes, and maybe I want to see a cute illustration of a baby peacock other times. It would be great to have a filter for 'real' vs 'artificial'. Perhaps ironically, using AI is the best way to make that distinction on my behalf.
karaterobot··on “Reading Rainbow” was created to combat summer reading slumps
We'd also have to figure out whether children who already loved reading watched Reading Rainbow, or if children who hated reading started liking it after watching. Since nobody has that data, I'll go with the aggregate.

> In other words, I don't believe the data you cited supports a conclusion that the show was ineffective at educating individual viewers.

I don't think it conclusively proves anything, but I do think it supports a skeptical position. The article doesn't cite anything supporting the notion that Reading Rainbow improved childhood literacy, so I'm wondering if you take the position that it did—and if so, on what basis?

karaterobot··on “Reading Rainbow” was created to combat summer reading slumps
They didn't start tracking in 1983, the numbers I linked start in 1971. The trend line is pretty much the same from 1971 to 1983 as it is from 1983 to 2006. In any case, a skeptical person would not look at that graph and say that there was a successful effort to improve childhood literacy represented on it.

It's true that we don't know the counterfactual: it's possible literacy would have plummeted precipitously starting in 1984 if Reading Rainbow hadn't been a bulwark. But I don't find that the most likely explanation, personally.

karaterobot··on What's happening to reading?
There may be a bit of that—watching things so that you can talk about them with other people. But, when I've done that (in the last 5 years or so), it's usually because there are one or two people I know who want to talk about a show. It's never been show that everyone I knew was watching.

The same is true of books, only moreso, since I only know a few other people who regularly read for pleasure, and with whom I might share a reading experience.

So it may be hyperbolic to say that no two people will be reading the same books, I think it's directionally true that we'll continue getting more and more narrowcasted, personalized content over time, thanks to AI, and devoting a larger portion of our time watching it, compared to content aimed at a wider audience.

karaterobot··on “Reading Rainbow” was created to combat summer reading slumps
This article lavishes well-deserved praise on the intentions behind Reading Rainbow. I know I loved the show as a kid.

But it seems like childhood reading scores were pretty much flat between 1983 and 2006, when the show was on the air: they only varied by 10-15 points on a 500 point scale[1], and there was no clear upward trend, it just sort of fluctuated. Reading for pleasure has never been lower among kids, either[2]. It doesn't seem to me that the mission of the show was achieved, if the mission was to make children read more books, and understand them more.

Ultimately I think it ended up just being a pleasurable way to have kids get distracted by a friendly, positive TV show. My guess is that if you want to improve reading scores and habits, parents have to do more than just turn the dial to PBS.

[1] https://www.nationsreportcard.gov/ltt/?age=9

[2] https://www.pewresearch.org/short-reads/2021/11/12/among-man...

karaterobot··on How I Use Kagi
Search engines are getting squeezed out by AI for me.

Kagi's search results are less polluted by SEO trash than Google, but there's still a non-zero amount of it. When I try to answer a question using Kagi (or any search engine these days) I end up feeling frustrated: there's so much information, and none of it is useful.

On the other hand, ChatGPT filters all SEO spam out for me, and typically does a decent job of answering my questions. I can follow the references it provides in its answer to verify what it says, and also learn more from external websites. It's a better user experience, with a better success rate for me.

Looking at my Kagi usage stats, I guess I'm not actually using it less from month to month (which I would have guessed I was). But, subjectively it still feels like I'm depending on it less for finding information on the web. I've given up on it for a lot of use cases, or it's no longer my first choice. I think my main use case is bang operators at this point, and that's where the numbers come from.

karaterobot··on What's happening to reading?
> In our current reading regime, summarized or altered texts are the exception, not the rule. But over the next decade or so, that polarity may well reverse

Most people don't read many books as it is. That's sad, but it's true for a couple decades. Those people will continue not reading.

Of the remainder, some people will continue to read books because reading books is enjoyable. It's not a chore, it's fun, so there is no reason to try to automate it away.

Another portion of people who read books only read them to get through them. Usually that's either because they want to understand some idea in them, or sometimes just so they can say they've read the book. These are the people who will use AI to summarize books rather than reading them line-by-line. I have zero problems with that, personally. That's not existentially troubling to me at all.

Nah, not too worried. Not about this anyway.

The thing that worries me about the future of literature is customization. I think having a shared monoculture—even a canon—is a good thing that we should strive for. We need it, to be part of the same community. The day that AI can easily just spin out a book that's exactly what you ask for, there won't be any reason for any two people to read the same books, let alone the same twenty million people. That will just accelerate the cultural fragmentation and attended alienation our culture has been living through for a while.

karaterobot··on "English Translators of Homer": A Review
> but it's not a full translation

If it's a translation at all, then we need a new word for when you change a text from one language into another language. I think 'retelling' is a perfectly good word to use here!

karaterobot··on When Figma starts designing us
> To me it’s categorically right: Auto Layout specifically is intended to restrict the possible layout options.

Except that you do not need to use Auto Layout if you don't want to mimic the flow of objects inside a flexbox container. You can just use a regular frame, and position things freely within it. Or, you can use an Auto Layout, but then absolutely position arbitrary elements inside that Auto Layout frame if you want. Auto Layout does not restrict a designer's layout options, it only adds to them.

karaterobot··on When Figma starts designing us
> A concrete example is Auto Layout... In practice, this locks the design in place and severely limits the possible expressions. You can’t drag things around freely or try odd combinations of layouts.

I'm a designer, I've used Figma since 2018, and this is incorrect. And not even incorrect in an "I feel differently, but I see what you mean" way. It's the opposite of correct. It's categorically wrong. Autolayout makes it easier to slap layouts together quickly, and change them quickly. The alternative is selecting the object and moving it with the arrows keys, which accomplishes the same thing but is slower, harder, less precise, and worse. It's not creatively empowering to manually align and space objects.

> “Ready for dev” implies that the creation is done and that the developer is merely there to execute the designer’s vision

Don't worry, no engineer I've ever worked with has shared your confusion.

"Ready for dev" is a work management trigger, like closing a ticket—or, more accurately, marking it ready for review. It doesn't mean anything except "this is ready for dev to look at and leave feedback". There is nothing about flagging a section as `ready for dev` that forces engineers to work on it as though it were canon law.

karaterobot··on Claude Code Pro Limit? Hack It While You Sleep
This project is about getting around the extremely aggressive throttling Claude applies.
karaterobot··on The Real GenAI Issue
This article read like a conspiracy theory, and it offers no evidence for its position. It's not even really making a case. It's just stating that companies are putting money behind AI so that they can lay off employees and reduce the quality of their products, and by saying that they apparently think they've made an argument.
karaterobot··on Lessons from 863 episodes of This American Life
Well, for any future archaeologists reading this, please know that This American Life was a great show that was made for a specific audience, by a specific set of creators, and it absolutely did not represent the breadth of life in America at the turn of the 21st century. This was a common mistake: thinking a rather small niche was universal because it's what you see. It led to a lot of surprises.
karaterobot··on I want to leave tech: what do I do?
> one of the reasons I find it hard to leave a high paying job is because "underpaid" has always been the best predictor of job toxicity.

Being 'underpaid' is different than being paid less, though. Describing a job as underpaid is, almost by definition, assuming a company is exploiting its workforce. It's not hard to believe such a company would also be a toxic environment in other ways. But if two companies pay different amounts, but they're both paying a fair salary, it doesn't necessarily imply anything about the company that pays less.

karaterobot··on Don’t use “click here” as link text (2001)
Keep doing whatever you've been doing for the last 24 years, it's fine.
karaterobot··on The role of the University is to resist AI
> The way this technology works means that generative AI applied to anything is a form of slopification, of turning things into slop. However, where AI is undoubtedly successful is as a shock doctrine, as a way to further precaritise workers and privatise services.

You can't say both that AI produces worse results, and that it will be used to manipulate the job market: savvy companies would outcompete by not adopting AI, and hiring up the victims of AI layoffs. If either of his statements is true, the other is false.

This whole article, man. I don't know where to start with it. It definitely reminds me of grad school. In a bad way.

The university has been on a glide path toward irrelevance for quite a while—long before AI was a going concern—and the humanities and social sciences, in particular, have been skimming the treetops since at least when I was in school at the turn of the century. The role of the university is to teach and do research. AI can be a tremendous asset for both of those, and it's not going away, so deal with that reality.

karaterobot··on Engineer creates ad block for the real world with augmented reality glasses
cmd-f shows one person in this thread bringing that name up, guess who it was.
karaterobot··on Engineer creates ad block for the real world with augmented reality glasses
I can understand you thinking it's sad (it is), but I really can't understand not thinking people would want it. Look at camera filters. There's even a "touch up" feature in Zoom. Of course people would want it. Nothing about the way people's revealed preferences would lead me to believe this wouldn't get used by a lot of people.
karaterobot··on MCP: An (Accidentally) Universal Plugin System
I don't understand your point. Some of those things were buzzwords, some were impossible dreams, some changed the way the web works completely. Are you just saying that the future is unknown?
karaterobot··on As AI kills search traffic, Google launches Offerwall to boost publisher revenue
Let's not forget that online publishing was dead on its feet before ChatGPT ever showed up. What really killed their revenue was zero barrier to entry, combined with social media monopolizing the attention of users. Every publisher fighting for smaller and smaller shares of attention with more and more outlets, leading to a race to the bottom.
karaterobot··on Transparent Ambition
> Apple's designers (and those of many other companies) come back to the idea of translucency giving order and imbuing personality. I cannot for the life of me understand where this idea comes from. When multiple layers of different imagery shine through each other, I am not helped by this. The user interface placed on a semi-transparent panel is not more effective because it is set against a smeared mess of colors, nor am I emotionally fulfilled as a human being for knowing that the mess comes from a treasured personal memory.

This article acts as though design choices which are sub-optimal from a purely informational perspective, but which add personality and attraction to a product, have not earned Apple hundreds of billions of dollars in the past 25 years. The fact that he cannot understand where the idea comes from is okay, since he's not a product, UI, or industrial designer for Apple: that's their job, not his. The deeper question of why people form attachments to things that look 'cool' but have lower performance in some areas is indeed mysterious. But relentlessly pursuing that phenomenon has worked for Apple pretty well so far.

Liquid Glass isn't immediately impressive to me either, but it will either succeed or fail not based on whether it is more effective or not, but whether people like it or not, and those are two different things.

karaterobot··on Using an $8 smart outlet to avoid brainrot
Mullvad VPN's client blocks that stuff as well. That filter prevents me from unknowingly clicking a link here on HN and getting twitter on me.
karaterobot··on The JAWS shark is public domain
> an artist should be able to enjoy the fruits of his work in his lifetime if he wants to

Not if you're an illustrator doing work for hire. It's not unreasonable or unusual for the company who commissioned the art to own the copyright. It doesn't always work that way, but there's no reason to think Kastel was robbed without us knowing the actual terms of his contract with Universal. I assume he sold the copyright to Universal, and Universal fumbled the copyright after that, but that doesn't mean it reverts back to Kastel.

karaterobot··on Extracting memorized pieces of books from open-weight language models
Where in the article do the authors say this puts anything to rest? Here is their conclusion:

> Our results complicate current disputes over copyright infringement, both by rejecting easy claims made by both sides about how models work and by demonstrating that there is no single answer to the question of how much a model memorizes

I wonder if this is the sort of article that people will claim supports their side, even that it ends the debate without a knockout blow to the other side, when the actual article itself isn't making any such claim.

I'm sure you read the entire article before commenting, but I would strongly recommend everyone else does as well.

karaterobot··on The Zed Debugger Is Here
Not at all!
karaterobot··on The Zed Debugger Is Here
It's an unexpected position to take, though. You said you had been interested in Zed until they integrated AI. The response was "the AI is completely optional", which I'd expect would make you more likely to use Zed, since it removes your objection. But it doesn't change your position at all, which makes me suspect it's not that you're worried the AI would interfere with your workflow, but that it's there at all. So, is your position that the very fact that Zed allowed AI to touch it has infected it in some way?
karaterobot··on Andrej Karpathy: Software in the era of AI [video]
It's odd! The guidelines don't say anything about having to read or watch what the posts linked to, all they say is it's inappropriate to accuse someone you're responding to of not having done so.

There is a community expectation that people will know what they're talking about before posting, and in most cases that means having read the article. At the same time, I suspect that in many cases a lot of people commenting have not actually read the thing they're nominally commenting on, and they get away with it because the people upvoting them haven't either.

However, I think it's a good idea to do so, at least to make a top-level comment on an article. If you're just responding to someone else's comment, I don't think it's as necessary. But to stand up and make a statement about something you know nothing about seems buffoonish and would not, in general, elevate the level of discussion.

karaterobot··on Ask HN: How do I give back to people helped me when I was young and had nothing?
When I was young, one of the things my mentor taught me is that the older people get, the more they think about their legacy. They want to believe they made a lasting difference. So, one of the best gifts you can give them is to say "you made a difference, what you taught me stays with me all the time, and you would never believe me if I told you how many people besides me have benefited from it."
karaterobot··on The librarian immediately attempts to sell you a vuvuzela
I agree they'd love to do that in theory, and it seems technically feasible. What gives me hope on that front is that marketers and advertisers (let alone the companies that pay them) have never shown the slightest capacity for that level of subtlety. The most sophisticated adtech today, produced by networks of massive data collection and analysis, ultimately just tries to shove as many loud, disruptive ads in your face as possible.

I think if you had this incredible technology that could manipulate language to nudge readers in the softest possible way toward thinking a little bit more about buying some product, so that in aggregate you'd increase sales in a measurable way that nobody would ever notice, it would just quickly just devolve into companies demanding the phrase "BUY MORE REYNOLDS GARBAGE BAGS!!!!!!!!" at least 7 times.

karaterobot··on How to post when no one is reading
This article seems like it's telling you to keep writing even though nobody is reading right now, because some day they might, and you've got to be ready for that day. I think writing is worth doing for its own sake, or not at all. The economics of writing these days are such that I would never tell someone to keep doing it with the hope that one day it will pay off. Instead, write because you have to say something, or because you need to exercise your critical thinking, researching, planning, or just your follow-through, but not because you think you might get rich and famous for it.
← PreviousPage 2 of 34Next →