@eloquence@social.coop
@eloquence@social.coop avatar

eloquence

@eloquence@social.coop

Professional:
Engineering, https://freedom.press

Personal:
https://lib.reviews and other free/open projects.

Opinions my own :-)

This profile is from a federated server and may be incomplete. Browse more on the original instance.

eloquence, to random
@eloquence@social.coop avatar

One nice thing I have to sometimes remind myself on the fediverse --

If I see more posts by someone, it's not because an algorithm is trying to optimize my engagement, it's because they happen to be around and active while I am.

Which honestly makes me feel more connected. :)

eloquence, to random
@eloquence@social.coop avatar

I don't have a problem with passwords, I have a problem with terrible password strength checkers and mandatory password changes after 3 months.

eloquence, to random
@eloquence@social.coop avatar

Now Elon Musk is spreading FUD about Signal. If you needed another vote of confidence that Signal (end-to-end encrypted messenger maintained by a nonprofit) is the real deal, there you have it:

https://signal.org/download/

eloquence, to random
@eloquence@social.coop avatar

I've contributed a fair bit to free content (CC licenses) & open source projects over the years.

Personally, I want "my" stuff to be used to make AI models better. I use open licenses precisely because I want people to come up with interesting & hopefully beneficial new uses.

I understand why lots of folks feel differently, of course.

However, it's not a clear-cut legal situation, either. Training != inference; it's only model output that violates licenses that's unambiguously infringing.

eloquence,
@eloquence@social.coop avatar

Creative Commons itself takes the, IMO very reasonable, view that training AI models on copyrighted works constitutes fair use:
https://creativecommons.org/2023/08/18/understanding-cc-licenses-and-generative-ai/

The folks who are calling such training "theft" might regret what they seem to be implicitly asking for, i.e. much stricter copyright. Copyright law won't prevent Microsoft, Google, OpenAI or Adobe from making shady licensing deals, but they'll prevent the free/open community from keeping up.

eloquence,
@eloquence@social.coop avatar

Some of the anti- backlash seems to go hand in hand with an explicit or implicit defense and support for copyright -- a questionable institution that aggregates power with the Disneys & Apples of this world.

I am very skeptical that a just world is one that still makes heavy use of intellectual monopoly rights to secure individual incomes.

Copyright should, IMO, at best be regarded as a necessary evil, one which we have failed to rid ourselves of along with capitalism.

eloquence,
@eloquence@social.coop avatar

@nicol

That question ultimately will be settled in the courts, but there are certainly plenty of experts and civil society groups who wholly or partially agree with CC's interpretation.This submission by Pamela Samuelson (Professor @ Berkeley Center for Law & Technology and a widely recognized expert on fair use) et al. to the copyright office is worth a read:

https://www.regulations.gov/comment/COLC-2023-0006-8854

eloquence,
@eloquence@social.coop avatar

@nicol

The gist is that if the model in its "ordinary and routine operation" does not copy the original data it was trained on, the training is (in their opinion) likely to be fair use.

We'll see what the courts find. IANAL, but my impression is that OpenAI and others are likely to prevail on most counts. There are some categories where genAI is very likely to produce infringing outputs (e.g., ask for green ogres, get Shrek), which seems the kind of thing most likely to bite them in court.

eloquence,
@eloquence@social.coop avatar

@nicol

> surely it’s the only defense the creative workers of the world have?

I think framing creativity it in terms of defense/offense against Disney is playing into the cycles of exploitation that exist today.

In a world without copyright, Disney in its current form could not exist -- creating an entirely different playing field for creatives. And anyone could build on the world's entire body of creativity in new & exciting ways all the time, without worrying about "stealing" information.

eloquence,
@eloquence@social.coop avatar

@nicol

Disney's version of Winnie the Pooh is still "protected" by copyright and trademarks:

https://ipwatchdog.com/2022/02/22/public-public-domain-winnie-pooh-illustrates-copyright-limitations-public-domain-works/id=146207/

Look to Frankenstein, Dracula, Jane Austen novels, Sherlock Holmes, Arabian nights, or Greek mythology for the rich universe of creative works that the public domain enables.

Of course, all those derivative works are again "protected" by copyright.

eloquence,
@eloquence@social.coop avatar

@nicol

I disagree with calling copyright a "worker protection". If it was, it would not be something that can be hoarded into large piles that confer massive economic power.

Copyright is an intellectual monopoly right that ostensibly exists to incentivize creative works. And it's often the first thing creative workers have to sign away as part of employment agreements or contracts.

eloquence, to llm
@eloquence@social.coop avatar

What happens when China builds an ? DeepSeek just released v2 of its model, which is open source.

I tried it on deepseek.com. Ask it about Tiananmen square, and the chatbot self-censors its answer while it is generating (that presumably is limited to their deployment). On variations not caught by the filter, it refuses -- and replies in Chinese:

"The content of your question is not in line with the core values ​​of socialism, nor is it in line with China's laws, regulations and policies."

eloquence,
@eloquence@social.coop avatar

Ask it about the Uyghur, and it praises the equal rights and opportunities for all ethnic groups in China.

Ask it about criticisms of the Chinese political system, and it has none.

Ask it about criticisms of the American system, it has plenty.

Ask it to compare the two systems' advantages and disadvantages, it starts writing about America .. and then censors its entire answer as the filter detects it's about to say potentially critical things about China.

eloquence, to ai
@eloquence@social.coop avatar

Perplexity 's "Discover" feature includes LLM-generated news summaries from many different sources.

Similar summary services will likely become more widespread and more sophisticated.

From a reader's perspective, it has significant benefits -- consistent format, no ads, pulling from many sources (including potentially multilingual ones), etc.

But for news media, if readers don't click through to original sources, it creates further strain on business models (ads, donations, etc.).

eloquence, to random
@eloquence@social.coop avatar

A lot of people have commented on All New Atlas' horror movie style movement patterns, but for me what stands out the most is the resemblance to "The One Electronic", the ageless, prophetic robot from Evan Dahm's "Rice Boy" comic.

"The One Electronic", a prophetic, ageless robot from Evan Dahm's Rice Boy comics. He's dressed in a ragged purple/red coat, and his circular faceplate displays abstract-looking black and white art. Illustration from https://topatoco.com/products/ed-oneelectronic

eloquence, to godot
@eloquence@social.coop avatar

"Slay the Spire 2" being developed with is a huge milestone for and .

"Slay the Spire" was an indie mega-hit -- it currently has 130,000 "overwhelmingly positive" reviews on Steam. I've put some 180 hours into it -- that's how good it is.

The devs switched away from the proprietary engine due to enshittification. They, and other indie devs who have switched to open source engines, won't be coming back. This is how we win.

https://www.youtube.com/watch?v=krDFltgjLtE

eloquence, to random
@eloquence@social.coop avatar

Just finished "The Excavation of Hob's Barrow", a wonderful adventure. Set in 19th century rural England, you play as a researcher investigating ancient burial mounds. One such barrow, in the small town of Bewlay, may harbor secrets of great personal significance to our protagonist ..

Atmospheric music & pixel art and stellar voice acting, especially by Samantha Béart (Thomasina). You're unlikely to need a walkthrough -- puzzles are very straightforward.

https://cadgames.weebly.com/the-excavation-of-hobs-barrow.html

Protagonist Thomasina Bateman doesn't take crap from anybody. (In this animation, she is seen outside an inn in the rain. She carefully puts down her suitcase, then slaps a drunken patron, then picks up her suitcase again.)

eloquence, to python
@eloquence@social.coop avatar

@dangerzone is an open source app that uses container sandboxes to create a "virtual photocopy" of a document.

This is to protect journalists and others from malware & trackers that might be implanted in documents sent to them by someone pretending to be a confidential source.

We (@freedomofpress) are hiring a software developer on a 6 month contract to work on the project. Remote, $70/hour, >=3 years of Python dev experience.

More info here:
https://grnh.se/1d1255f45us

eloquence, to random
@eloquence@social.coop avatar

IMO one of the worst SaaS practices is locking higher seat counts under massively more expensive plans (looking at you Mailchimp). It's a great way to incentivize bad security practices for small orgs.

Role-based access control and the option to add seats should never be a "premium" feature. It's a basic security/privacy measure.

eloquence, to random
@eloquence@social.coop avatar

Nonprofit/open source job alert: @OpenArchive is hiring a Lead Mobile Dev. The org is focused on enabling human rights defenders to "share, archive, verify, and encrypt their evidentiary media."

1 year contract; hourly rate in the $80-100 range. Fully remote, overlap with US time zones is preferred. Senior role, with 7+ years experience listed under preferred qualifications.

https://open-archive.org/jobs/lead-developer

john_fisherman, (edited ) to random
@john_fisherman@mastodon.social avatar

Hey folks, I'm writing a piece for @fumacapt on why RSS matters and why it's cool.

Can you share your personal perspective, resources you go back to, clients for different platforms?

That would be ace.

It's going to be out next week, and it's in Portuguese.

eloquence,
@eloquence@social.coop avatar

@john_fisherman

I'm a big fan of NewsBlur. It's open source, but they run a hosted version as well.

I use both the web UI and their Android app which gives me basically everything I want -- being able to track read/unread stories across devices. My only criticism is that the UI is a bit clunky, but I've gotten used to it.

eloquence, to VisualNovels
@eloquence@social.coop avatar

If you enjoy and , check out the latest issue of , a free zine dedicated to the genre:

https://choicebeat.wordpress.com/2024/02/23/choicebeat-10/

It's chock-full of reviews; I contributed two:

  • Night Cascades, a mystery romance set in a New Age infused version of the 1980s, by @hanakogames

  • Seers Isle, an interactive graphic novel by @novabox in a fantasy medieval northern Europe setting. Themes include individual survival, and the clash between monotheism & polytheism

cheeaun, to random
@cheeaun@mastodon.social avatar

This is probably too hacky but can't help trying it out. Phanpy's shortcuts settings are saved as (private) notes on my own profile, which will allow export/import in the "cloud".

It's possible to use the API to add notes for other profiles, and they're hidden from (self, currently-logged-in) own profile. I'm using it to store data 🤪

NOT available on the dev site yet, still on my local machine.

Demo of the "cloud" import/export feature for Shortcuts Settings on Phanpy.

eloquence,
@eloquence@social.coop avatar

@nikclayton @cheeaun @thisismissem

IMO any privacy consideration for storing per-app settings on the server could be mitigated by making this clear in the permission dialog.

"This app may store settings, drafts, or other information on this server" or similar.

I don't know if the idea is feasible in terms of separating different apps' settings storage. It seems preferable to me over a third party service though. I'd much rather have my instance store such stuff than, e.g., Google.

eloquence,
@eloquence@social.coop avatar

@nikclayton @cheeaun @thisismissem

But I also agree with Em's comment on the issue. Limiting the storage size is probably the easiest way to mitigate those concerns.

That would take a draft feature off the table, but I think drafts are a good fit for localstorage anyway. It's always a bit problematic to send unfinished text to a server potentially without the user realizing it.

eloquence, to random
@eloquence@social.coop avatar

Disappointing to see downsize mozilla.social before they even properly got it off the ground. Lack of perseverance will doom any effort to achieve transformative impact.

https://techcrunch.com/2024/02/13/mozilla-downsizes-as-it-refocuses-on-firefox-and-ai-read-the-memo/

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • provamag3
  • kavyap
  • DreamBathrooms
  • InstantRegret
  • magazineikmin
  • thenastyranch
  • ngwrru68w68
  • Youngstown
  • everett
  • slotface
  • rosin
  • ethstaker
  • Durango
  • GTA5RPClips
  • megavids
  • cubers
  • modclub
  • mdbf
  • khanakhh
  • vwfavf
  • osvaldo12
  • cisconetworking
  • tester
  • Leos
  • tacticalgear
  • anitta
  • normalnudes
  • JUstTest
  • All magazines
    •