A simple question. I know LLMs are currently not a replacement for authors. Will that remain true in 5 to 10 years?

EDIT: No. I never expected to earn a living either mostly or exclusively by selling books. There are however many "side gigs" in my country that can greatly benefit from being published by a real company. Ultimately though, I'm not in it primarily for the money. But I wonder what the future holds for fiction as a whole.

This is really just a follow-up update to my old post, Did NPR just shut down support for RSS?, but that post is a week old and I wanted to make sure this update gets eyes on it.

I heard back from NPR this morning, and they indicated that this was not a policy decision, but an issue that they have now fixed. NPR RSS links once again work outside of the US/Canada.

Passingly curious how this kind of thing happens accidentally ... any informed guesses? My best guess is that they intentionally geoblocked something else, and it accidentally got extended to the RSS.

I stay current on tech for both personal and professional reasons but I also really hate hype. As a result I've been skeptical of AI claims throughout the historic hype cycle we're currently in. Note that I'm using AI here as shorthand for frontier LLMs.

So I'm sort of a late adopter when it comes to LLMs. At each new generation of models I've spent enough time playing with them to feel like I understand where the technology is and can speak about its viability for different applications. But I haven't really incorporated it into my own work/life in any serious way.

That changed recently when I decided to lean all the way in to agent assisted coding for a project after getting some impressive boilerplate out of one of the leading models (I don't remember which one). That AI can do a competent job on basic coding tasks like writing boilerplate code is nothing new, and that wasn't the part that impressed me. What impressed me was the process, especially the degree to which it modified its behavior in practical ways based on feedback. In previous tests it was a lot harder to get the model to go against patterns that featured heavily in the training data, and then get it to stay true to the new patterns for the rest of the session. That's not true anymore.

Long story short, add me to the long list of people whose minds have been blown by coding agents. You can find plenty of articles and posts about what that process looks like so I won't rehash all the details. I'll only say that the comparisons to having your own dedicated junior or intern who is at once highly educated and dumb are apt. Maybe an even better comparison would be to having a team of tireless, emotionless, junior developers willing to respond to your requests at warp speed 24/7 for the price of 1/100th of one developer. You need the team comparison to capture the speed.

You've probably read, or experienced, that AI is good at basic tasks, boilerplate, writing tests, finding bugs and so on. And that it gets progressively worse as things get more complicated and the LoCs start to stack up. That's all true but one part that has changed, in more recent models, is the definition of "basic".

The bit that's difficult to articulate, and I think leads to the "having a nearly free assistant" comparisons, is what it feels like to have AI as a coding companion. I'm not going to try to capture it here, I'll just say it's remarkable.

The usual caveats apply, if you rely on agents to do extensive coding, or handle complex problems, you'll end up regretting it unless you go over every line with a magnifying glass. They will cheerfully introduce subtle bugs that are hard to catch and harder to fix when you finally do stumble across them. And that's assuming they can do the thing you're asking then to do at all. Beyond the basics they still abjectly fail a lot of the time. They'll write humorously bad code, they'll break unrelated code for no apparent reason, they'll freak out and get stuck in loops (that one suprised me in 2025). We're still a long way from agents that can actually write software on their own, despite the hype.

But wow, it's liberating to have an assistant that can do 100's of basic tasks you'd rather not be distracted by, answer questions accurately and knowledgeably, scan and report clearly about code, find bugs you might have missed and otherwise soften the edges of countless engineering pain points. And brainstorming! A pseudo-intelligent partner with an incomprehensibly wide knowledge base and unparalled pattern matching abilities is guaranteed to surface things you wouldn't have considered.

AI coding agents are no joke.

I still agree with the perspectives of many skeptics. Execs and middle managers are still out of their minds when they convince themselves that they can fire 90% of their teams and just have a few seniors do all the work with AI. I will read gleefully about the failures of that strategy over the coming months and years. The failure of their short sightedness and the cost to their organizations won't make up for the human cost of their decisions, but at least there will be consequences.

When it comes to AI in general I have all the mixed feelings. As an artist, I feel the weight of what AI is doing, and will do, to creative work. As a human I'm concerned about AI becoming another tool to funnel ever more wealth to the top. I'm concerned about it ruining the livelihoods of huge swaths of people living in places where there aren't systems that can handle the load of taking care of them. Or aren't even really designed to try. There are a lot of legitimate dystopian outcomes to be worried about.

Despite all that, actually using the technology is pretty exciting, which is the ultimate point of this post: What's your experience? Are you using agents for coding in practical ways? What works and what doesn't? What's your setup? What does it feel like? What do you love/hate about it?

I didn't watch the whole video and I'm not familiar with the channel so I don't want to make this a link post, but here's the source: The Lunduke Journal

I watched up to the point where the author explains how Microsoft tends to turn on all the privacy invading settings every time they push an update (not surprising). I guess if I had to use Microsoft products, I'd try to disable automatic updates and just do them twice a year in one go, while also turning off the settings I want off. Would it be practically feasible? I don't know. Having to go to those lengths to use some software just seems ridiculous.

I'm getting a little bit worried these days about the tracking features in chrome and was looking into Zen Browser. Does anyone here use it? Any thoughts?
Also looking for any simple non tracking browser alternatives!

I'd say that on Tildes as on other platforms, you see a lot more posts responding to a topic than you see comments on existing posts.

I get it. Responding to a prompt with a thoughtful, top-level comment is expressive, can often be therapeutic and comes with the bonus of possible comments by others on your entry.

Comments on existing thoughts are less sexy and possibly less fulfilling because you're riffing off of another person's idea, but as a reader and a community member, seeing user to user interaction is the best part of a social network.

I'm a perennial commenter - at best because I love conversation, at worst, with the hope that I can digress from the mainline conversation.

Where do you fall?

*Edit: I've just learned the difference in terminology between a top level comment and a comment. Edited to avoid confusion.

I had an interesting experience this week. Not all union meetings are interesting, even if they are useful. Yesterday was a pleasant exception where it was both useful and interesting. For the first time, I witnessed AI coming up as a topic of conversation. There is no secret that people fear losing their jobs due to AI automation, and sure enough I saw proof of it to the extent that the union may consider adding some clauses around protecting jobs from AI.

How is it at your workplace? Where I work, this year I witnessed a very strong push to use AI. Messaging around using AI at town halls, messaging around using AI in team meetings, articles on the intranet site, IT events around how to craft good prompts, etc. I would not be surprised if they tied some leaders' bonuses to how much they can get their teams to use AI. This part is quite annoying to me, not to mention deceitful. If I were a leader I'd straight up tell my team about it. I am not a leader - leaders are not part of the union to begin with.

The whole thing made me also think about how my colleagues use AI. It really is a mixed bag. I see everything from the person who runs a 2-line email through AI five times to finetune every word, to myself who only reach for AI when I am stuck and it's just much faster than a search engine/forums/videos to solve my issues (for example needing a script in a programming language I am not familiar with).

I was trying to find out why Lidarr wasn't matching my copy of The Cure's Greatest Hits. Found out I've got some bootleg Russian release that's catalogued on discogs (I eventually found the musicbrainz release and updated my profile to include bootlegs). So I search "Lidarr use specific discogs release" and the duck duck go search assist spat out some text about Lidarr not using discogs and cited this Tildes post.

It's curious because that post is 3yrs old and doesn't talk about discogs integration in Lidarr, just one mention of discogs in the post and some folks talking about Lidarr in the comments (It did cite a relevant GitHub issue about it though). The AI response mentioned that some users track new releases with Lidarr and downloads disabled, while covered in the post, it seems fairly tangential to my query.

I'm curious why it decided to check or cite a tildes post. No tildes posts came up in the first couple pages of search results. I use tildes from the same location, though on my phone where this query was on my desktop, and have done a couple DDG queries using "site:tildes.net" on my phone.

Has anyone else seen a search assist cite an unexpected site? Not unexpected as in irrelevant, that's all too common, but small and specific sources.

I have a Starlabs Labtop in great shape that I would like to sell. Problem is, I can't garner much interest on Craigslist/Facebook Marketplace. Are there any computer Buy/Sell sites in Canada such as Pinkbike.com for used bicycles?

I use a Mac for work, Linux for my personal machine, and the same mouse between the two. I’m accustomed to the dpi/acceleration on Mac, so when I switch to my PC at the end of the day there are slight inaccuracies with the pointer. My wrist compensates to cover the gaps and I develop wrist strain after short periods of use.

Does anyone know how to make the Linux mouse dpi/acceleration as close to identical as possible with my Mac?

chimera linux w/ gnome btw

I've been trying to get a more rounded understanding of the impacts that "AI" has had since ChatGPT went viral back in 2022.

I've found it easy to gather a list of negative impacts, but have struggled to point to many positives.

I was curious if there were folks who have used any of these AI tools, and would willing to share any positive impacts those tools have had in their lives. I'm particularly interested in the text, audio, image, and video generation tools that have appeared since ChatGPT went viral, but please share anything else that you think fits.

I have resubscribed to Netflix after cancelling about 6 months ago and I think there's been a change to the GUI which I really don't care for. I could be wrong but didn't the TV app use to work in the following way: after selecting a movie or television show, the first click would open a menu that would then allow further options including playing the movie or an episode of the TV show? I miss this. Currently, they're is one option at the to of my screen when I open the app that has a selectable 'More Info' option but everything else does not and the media just starts playing immediately upon selection.

I miss the 'menu on first click' behavior. Is there a way to get back to that?

Thanks.

I keep seeing this term but I have no idea what it means ?

and what does it actually mean in practice. for example, what it can actually do, it seems to only be used for algorithms and such. not personal computing. I assume I don't understand because I'm unfamiliar with Quantum mechanics

Hello!

I recently reworked my setup such that I have my drive which holds the operating systems (dual boot of Windows 11 and Linux Mint Cinnamon) and another drive which holds all my actual data. This is my first dive into trying to move to Linux as my daily driver but I'm noticing some performance issues.

The first thing I noticed is that transfer speeds are much less than expected. Copying a large file within my data drive I get about 300MB/s on Linux, which is pretty slow for M.2 drives- I get about 10x that speed doing the exact same operation on Windows. I could be okay with this but I also noticed that some video files, like the 4K mp4s off my phone, are virtually unplayable. They'll run at maybe 1fps and/or bog down the UI so much I cant even use the seek. This is the case in both the default media player and VLC. These same files play with no issue on Windows.

I suspect the reason for all this stems from my data drive being NTFS, though my file system and Linux knowledge is pretty weak so this is just a theory. Any ideas / best practices that might help me here?

I decided to spin off this topic based on the discussion in the Leaker reveals which Pixels are vulnerable to Cellebrite phone hacking.

Since I kinda by chance got rid of my dependency on Google Photos (as I installed Immich on my server which does the same but on my own hardware) I decided I want to try and de-Google my life more.

I was thinking about using my own domain to send e-mail from, but since I don't run e-mail server myself and I don't even want to - because of security reasons from multiple directions - I would like to use some paid e-mail service provider that would host the e-mail for me and I would append my domain to it.

I probably know technicalities on how to do that - through setting DNS MX record on my domain provider and pointing it to my e-mail provider. But I don't know which provider to choose.

I would like to sync my contacts (not through import/export feature via .vcf file, rather automatic synchronization like Google/Microsoft/Apple accounts do), I would like to sync calendar and have the ability to share it with other people (can be another account on the service if not outsiders) and of course e-mail, preferably one that doesn't rely on their own app so I can use ie. Thunderbird.

I would like to kindly ask anyone here in audience to share their experience with their providers, if you use some. It would be great if you used the service as I would do or if you could try my proposed usage on your account.

Thank you all who respond to me. I may have questions if you do :-)

I know about Fastmail which is appealing for not being US (things are kinda unsettling there right now), for having contacts, calendar and plenty space, but isn't exactly cheap - not expensive though. I know about Tuta, which is cheaper for 1-3 accounts but not for 4 and above, it has less space (still enough) but I don't know about contacts there at the moment. I found this Wikipedia comparison page and had a look there, but I would like to hear personal experiences and thoughts.