ell1e

joined 1 year ago
sorted by: new top controversial old
Do you write all your code by hand? Join the rest of us in this public list in c/fuck_ai@lemmy.world
[–] ell1e@leminal.space 1 points 5 hours ago

Perhaps submit yourself anyway and just put that explanation into the pull request. I don't think the list is designed to be super bullet proof, I was merged without directly linking any code too (although I suppose it's not too hard to find). At the end of the day, a lot of the internet runs on the honor system still, for better and for worse.

Do you write all your code by hand? Join the rest of us in this public list in c/fuck_ai@lemmy.world
[–] ell1e@leminal.space 2 points 5 hours ago* (last edited 5 hours ago)

Kind of funny (or sad?) that lemmy itself probably wouldn't make it on the list, given they seem adamant on allowing LLM coding: https://github.com/LemmyNet/lemmy-docs/pull/414/changes

~~Curiously enough I wasn't able to actually find any lemmy commit marked as created with the help of AI, other than some bug located by AI without indication that an AI fix was used. I wonder if that means either they don't adhere to their own rule, or whether they're not actually using AI but just felt like really being pro-AI anyway. Why though? I'm so curious.~~

Update: seems like they don't put it into the commit log, just the merge request info: https://github.com/LemmyNet/lemmy/pull/6304

‘Open-Washing’ Is Everywhere in AI. Four Criteria Cut Through It. in c/opensource@programming.dev
[–] ell1e@leminal.space 1 points 8 hours ago* (last edited 7 hours ago)

So how would that work given LLMs apparently can't do much other than rearrange/plagiarize the training data?

I think that's where the article writer's definition comes from. A model probably won't seem very open to most people if the output licensing isn't really compatible with free use.

‘Open-Washing’ Is Everywhere in AI. Four Criteria Cut Through It. in c/opensource@programming.dev
[–] ell1e@leminal.space 3 points 12 hours ago

Is that meant to be some sort of humorous remark regarding the OpenAI name? I think most observers would agree that the name isn't particularly fitting, so that seems like a fair observation.

‘Open-Washing’ Is Everywhere in AI. Four Criteria Cut Through It. in c/opensource@programming.dev
[–] ell1e@leminal.space 4 points 13 hours ago* (last edited 12 hours ago) (2 children)

The "open" criteria listed in the article seem applied inconsistently by the writer:

In our recent AI and Ethics article, "The open-source advantage in large language models (LLMs)," my co-authors and I operationalize this even more concretely: a model qualifies as open-source only when its architecture, training code, model weights, and training data are all publicly available under licenses that permit unrestricted use, modification, and redistribution. By that standard, very few models qualify. The Allen Institute’s OLMo, EleutherAI’s GPT-NeoX, and LLM360’s K2 are among the handful that meet these four criteria.

I just checked the first one, OLMo. It just ingests random web pages, as far as I can tell. How would that possibly qualify for "training data [...] available under licenses that permit unrestricted use"? Why does the article writer think it would?

How would anything but pure CC0 training data fit that, given attribution requirements are so common and I assume they're a restriction on use (that most models ignore)?

The article apparently not wanting to admit this, and/or not having done the basic research to check, makes it seem as bad to me as the open-washing it is complaining about.

(However, perhaps I'm misunderstanding something here?)

Codeberg Git Hosting Platform Bans Projects Mostly Written by Generative AI in c/opensource@programming.dev
[–] ell1e@leminal.space 12 points 4 days ago (1 children)

Yup. And looking at the fallout from the other comments, seems like it may have been a good idea to take a stand.

(I don't like people becoming upset, but clearly some people embrace AI a little much, and if you look at Forgejo for example then you'd know Codeberg was always on some level anti AI due to the ethics and all that. And as far as I can tell, Codeberg always wanted to be a somewhat opinionated pro ethics code host. Now I understand some people don't like where the line has been drawn, but it's not suprising Codeberg wants to draw a line somewhere.)

A nice Goldblum farewell in c/funny@lemmy.world
[–] ell1e@leminal.space 5 points 2 weeks ago (1 children)

Video doesn't seem to load for some reason. But the interview screenshot that you included is very funny!

I love the original Jurassic Park, sad news for sure.

Europe must choose between AI and climate goals, data center lobby says in c/fuck_ai@lemmy.world
[–] ell1e@leminal.space 8 points 2 weeks ago (1 children)

Was my thought as well. Like, how thankful of the billionaires to give us yet another good reason to dump them.

Anyone else mass-renaming env variables and instantly regretting it? in c/programming@programming.dev
[–] ell1e@leminal.space 8 points 2 weeks ago* (last edited 2 weeks ago)

I feel like it depends on the project.

Typically in the ones I work on, whenever I spot an inconsistency I fix it no matter how much of a mess that causes. Yes it hurts at first, but it may hurt more down the line when the weirdnesses accumulate to the point where it impacts operations and it'll be a much bigger problem to fix.

However, I can totally see a more chill strategy work better for a project that is mostly just relatively simple code, like perhaps some website deployments. In my opinion it depends a lot on whether the code is generally already complicated, which is when you'll typically want to refactor earlier than later.

So it begins in c/privacy@lemmy.ml
[–] ell1e@leminal.space 1 points 3 weeks ago* (last edited 3 weeks ago)

I disagree. A widely useed digital id in itself is dystopian due to the one-click deplatforming risk, and if you require age checks everywhere for regular discussion forums like Reddit then it'll be widely used.

UK Online Safety Act in the EU - starting now? in c/privacy@lemmy.ml
[–] ell1e@leminal.space 1 points 3 weeks ago (1 children)

Might not be much left to go to, if people everywhere give up on protesting and this is adopted widely. I recommend sending an email even if it seems pointless.

So it begins in c/privacy@lemmy.ml
[–] ell1e@leminal.space 11 points 3 weeks ago (2 children)

on paper I’m not against it

Perhaps you should be, since there doesn't seem to be a non-dystopian way to do age checks on the internet at a large scale (as in, for more than e.g. sites dedicated for porn, and other very narrow examples). See Cory Doctorow write about Age Verification of any kind: https://pluralistic.net/2025/08/14/bellovin/ Or look at this EU wallet writeup: https://gitlab.opencode.de/bmi/eudi-wallet/wallet-development-documentation-public/-/work_items/13

22
UK Online Safety Act in the EU - starting now? (leminal.space)
submitted 3 weeks ago* (last edited 3 weeks ago) by ell1e@leminal.space to c/privacy@lemmy.ml
3 comments fedilink
 

🚨 UK Online Safety Act in the EU - starting now? 🚨

It appears I had suspected right and big platforms will think the UK Online Safety Act equivalent for the EU has already been decided by the EU commission in July 2025, with a deadline of July 2026 (a few days). At least Reddit acts like it: https://support.reddithelp.com/hc/en-us/articles/50368431806484-European-Union-Digital-Services-Act-DSA

I talked about all of this here on lemmy, with more detailed explanations: https://leminal.space/post/31858818/21120139

How all of this went for the UK previously: https://www.theguardian.com/commentisfree/2025/aug/09/uk-online-safety-act-internet-censorship-world-following-suit https://www.aljazeera.com/opinions/2025/11/6/britain-calls-it-safety-it-is-censorship

I'm worried this might be the beginning of the end of anonymous public discourse in the EU...

For anybody reading in the EU, consider emailing one of your representatives and demand they undo this age verification surveillance stuff: https://fightchatcontrol.eu/#delegates (Site is unrelated, it just happens to have a list of representatives.)

For anybody in the US, find your representatives here: https://www.senate.gov/senators/senators-contact.htm Since the US has similar plans: https://act.eff.org/action/tell-congress-don-t-force-age-checks-online

68
A curious response (lemmy.nz)
 

A curious response on the lemmy bug tracker... https://github.com/LemmyNet/lemmy-docs/issues/413

Disclaimer: please don't go and harrass anybody, if you happen to reach out to anybody in any way please be calm and constructive.

37
GitLab: "The World we're building for: SLOPPPP!!!" (dramatization) (leminal.space)
submitted 2 months ago* (last edited 2 months ago) by ell1e@leminal.space to c/fuck_ai@lemmy.world
5 comments fedilink
 

The world we're building for [...] Software will be built by machines, directed by people. AI is the substrate on which future software gets built.

AKA

The world we're building for: SLOP

At least this is how this reads to me: https://about.gitlab.com/blog/gitlab-act-2/ They just released this some hours ago.

If anybody is self-hosting GitLab here, then have my condolences. A possible alternative might be Forgejo.

125
"We're repeating German history, apparently" (gitlab.opencode.de)
 

You might find this write-up on the EU age verification plans interesting. It links the actual EU plans with quotes from them, too.

18
Is this Raspberry Pi 5 damage fixable? How do you guys find good repair shops? (lemmy.nz)
 

(Sorry if I'm posting this in the wrong community!)

A Raspberry Pi 5 I use (not the one in the photo) fell off the table, and the black square component pointed at by the arrow was straight knocked off the board. Yeah, if you needed a reason why you should always use a case, this is one, ooops lol...

It looks like it came off cleanly, leaving two square metal pads below. I still have the component, the small black box. This happened while the Raspberry Pi was running, and the power LED went from green to red in an instant. Afterward, booting was no longer possible. So, my questions:

  1. If this happened while it was running, is it likely that this fried something else and it's not worth getting fixed? Or is this likely fixable?

  2. If I want to get this fixed, how do you guys find a good local repair shop? I heard some apparently do sloppy soldering jobs, and given the RAM prices I would like not to have this Raspberry Pi unnecessarily damaged further if it's salvageable. I live in Freiburg im Breisgau in Germany, if anybody knows good local places.

29
The Linux Foundation apparently thinks you can look at a snippet to tell if it's stolen (hachyderm.io)
 

The Linux Foundation apparently thinks you can look at a snippet real hard for a minute or so, and you'll magically figure out if it's stolen or plagiarized from somewhere. Otherwise, I don't understand how their policy would possibly work. It's not like they're linking some sort of plagiarism checker, and as far as I know, those aren't reliable anyway.

Does somebody understand how that's meant to work? I'm curious. If I'm judging them too harshly here, I would like to know.

84
Lemmy may be heading down the path of LLMs (leminal.space)
submitted 4 months ago* (last edited 4 months ago) by ell1e@leminal.space to c/fuck_ai@lemmy.world
54 comments fedilink
 

Sadly, it seems like Lemmy is going to integrate LLM code going forward: https://github.com/LemmyNet/lemmy/issues/6385 If you comment on the issue, please try to make sure it's a productive and thoughtful comment and not pure hate brigading.

Consider upvoting the issue to show community interest.

Edit: perhaps I should also mention this one here as a similar discussion: https://github.com/sashiko-dev/sashiko/issues/31 This one concerns the Linux kernel. I hope you'll forgive me this slight tangent, but more eyes could benefit this one too.

106
Should anybody trust Firefox again unless they put "we won't sell your data" back into the privacy policy? (Have they done so...? I can't tell.) (leminal.space)
submitted 4 months ago* (last edited 4 months ago) by ell1e@leminal.space to c/privacy@lemmy.ml
85 comments fedilink
 

Firefox is trying to gain back user trust with this video: https://www.youtube.com/watch?app=desktop&v=O-xyNkvIB9g

This is a legit question: Should anybody trust Firefox again unless they put "we won't sell your data" back into the privacy policy? I'm actually not sure if they haven't already done so, let me elaborate:

https://brave.com/privacy/browser/ Brave: "We do not sell, trade, or transfer your information to any third parties." This seems to obviously be in the legally binding text part. As is this one: "It’s Brave’s policy to not collect personal data1 unless it’s necessary to provide services to our users, or to meet certain legal obligations. We do not buy or sell personal data about consumers." (Disclaimer: I'm not a lawyer.)

However, for Firefox it seems ambiguous to me, which worries me: https://www.mozilla.org/en-US/privacy/firefox/#notice There is no appearance of "sell" in the entire privacy document, excpet for the top summary where i'm not sure if it's at all legally non-binding.

Does anybody know if it is legally binding? If Mozilla were serious about it, why would they leave it ambiguous whether it is...?

Based on that, I'm not sure if Mozilla's video about getting users back is worth trusting. I wonder if it's just me.

Update for clarification: I'm not using Brave myself, and this isn't a suggestion anybody should blindly do so.

18
Interesting video on why apparently moltbot and other AI agents are dangerous (youtu.be)
 

Interesting video on why apparently moltbot and other AI agents are dangerous. I'm not an expert but it seems quite concerning, especially the prompt injection. (I assume amplified by the issue with current LLMs apparently being unable to think logically: https://www.forbes.com/sites/corneliawalther/2025/06/09/intelligence-illusion-what-apples-ai-study-reveals-about-reasoning/ )

Sorry if this is considered off-topic or was posted before.

38
A sourced summary article why AI is bad (github.com)
 

Here's a sourced article that actually shows the hands-on seeming plagiarism of AIs, how common it is, how the logical reasoning seems to be lacking, and so on. I thought perhaps people here would find it useful to convince friends that are misled by the AI craze.

13
Question about community maintained free IP geo lists (leminal.space)
 

cross-posted from: https://leminal.space/post/24911246

I'll be self-hosting a service with user submissions soon, so I'm worried about the https://howto.geoblockthe.uk/ situation.

Based on this I've wondered, are there any community maintained geo block lists that might be useful? All database options I found are either 1. an on-demand online service which seems questionable for privacy reasons, or 2. IPv4 only, or 3. have weird terms of use with a gag clause regarding the entire company making it and other weird stuff.

I'm not a fan of geo blocking in general, but the situation is what it is.

PS: Please don't discuss the Online Safety Act itself too much in the comments, or whether somebody should be using a geo ip to handle this. While I might appreciate useful input on that, I'm hoping this post can remain a resource for those who are looking for such a database for other reasons as well.

9
Community maintained free IP geo lists (leminal.space)
 

cross-posted from: https://leminal.space/post/24911246

I'll be self-hosting a service with user submissions soon, so I'm worried about the https://howto.geoblockthe.uk/ situation.

Based on this I've wondered, are there any community maintained geo block lists that might be useful? All database options I found are either 1. an on-demand online service which seems questionable for privacy reasons, or 2. IPv4 only, or 3. have weird terms of use with a gag clause regarding the entire company making it and other weird stuff.

I'm not a fan of geo blocking in general, but the situation is what it is.

PS: Please don't discuss the Online Safety Act itself too much in the comments, or whether somebody should be using a geo ip to handle this. While I might appreciate useful input on that, I'm hoping this post can remain a resource for those who are looking for such a database for other reasons as well.

view more: next ›