Mastodon Feed: Posts

db@social.lol ("David Bushell 🪿") wrote:

interesting approach to light/dark theme toggles:
https://vale.rocks/micros/20260810-0330

accounts for the system theme state too with only two buttons

jwz wrote:

Youtube's joke of a fair-use appeal process, Total Eclipse of the Content ID Robots Edition.

"Good news! Your dispute wasn't reviewed within 30 days, so the claim on your YouTube video has now been released." Bonnie Tyler died last month, and so the...
https://jwz.org/b/yk-Y

glyph ("Glyph") wrote:

RE: https://mastodon.social/@glyph/117069160758816522

NB, if you did not already know this: Y2K was a very serious problem and a huge number of people worked extremely hard to make sure that it did not result in a civilizational crisis. The Best Buy sticker meme is indeed funny but that's because it lampoons the public's poor understanding and the industry's poor communication about what the problem was, not because the problem was not real and hugely risky.

Boosted by isagalaev ("Ivan Sagalaev :flag_wbw:"):
limebar ("Lime Bar") wrote:

I am a little shocked at the lack of solidarity I see between artists working in different mediums wrt genAI.

Musicians who would never use it to make music use it to make cover art

Graphic artists reach for it to make music over their videos

Writers use it to illustrate their articles

And, considering software development an art, many use it to create web sites or apps to host their art.

None of this is necessary. All of this is harmful to the other artists.

Collaborate ffs.

Stop it.

glyph ("Glyph") wrote:

The OP I'm quoting here is not especially guilty of this, look downthread a few posts and you can see they're quickly following up with specific resources to check out. And I'm not a paragon of virtue here either, I am exasperated plenty often. But the point I was trying to make here is that we need to acknowledge the effort and make these type of resources more front and center when we write about this problem, try not to let them be afterthoughts. Lead with "Here's where you find some cake:"

glyph ("Glyph") wrote:

A little clarification here: I'm not saying that solidarity is not *worth it*. Obviously, you _gotta_ do it. It's arguably the whole point.

What I was trying to say here is that when you say to someone who is considering using some AI to avoid this type of community building, expressing exasperation is insufficient. "Why don't you just experience some genuine human connection?" is the "let them eat cake" of the loneliness crisis.

"Use ChatGPT" is still "let them eat diarrhea lettuce" though

isagalaev ("Ivan Sagalaev :flag_wbw:") wrote:

RE: https://newsie.social/@TheConversationUS/117068751158866830

Hey, you misspelled the last word, there shouldn't be an 'f' in it!

neatnik@social.lol ("Neatnik :prami:") wrote:

The embroidery art on the new Nidoran♀ Pokémon card is really something. This is the debut card illustration by the embroidery artist tayu (not counting their prior illustration contest submission).

neatnik@social.lol ("Neatnik :prami:") wrote:

jsonstein@masto.deoan.org ("Jeff Sonstein") wrote:

genius https://www.pandora.com/artist/hank-mobley/another-workout/hanks-other-soul-remastered-2006/TRqmfP9Kxrw3KjV?part=ug-desktop&corr=8104885

Boosted by glyph ("Glyph"):
ehashman@cloudisland.nz ("e. hashman 🇵🇸") wrote:

New blog post: some bash shenanigans for managing Python virtualenvs https://hashman.ca/managing-venvs/

glyph ("Glyph") wrote:

("wrong" in that the media believes that climate change is real)

Boosted by jsonstein@masto.deoan.org ("Jeff Sonstein"):
vicfroh ("Victator") wrote:

#usempiredecline #Rome #fallofrome #christianity #SeperateChurchAndState #oligarchy

glyph ("Glyph") wrote:

technically speaking, this is actually secondary gell-mann amnesia. primary gell-mann amnesia is the phenomenon where everyone forgets that michael crichton coined this term in a speech explaining that "the media" is wrong about climate change, because sometimes their reporting on cutting-edge theoretical physics is oversimplified to the point of being wrong

Boosted by jsonstein@masto.deoan.org ("Jeff Sonstein"):
BBC@c.im ("BBC News") wrote:

#BBC #News Evidence that South African special forces murdered top detective shared with BBC https://w.st/JrML9

Boosted by jsonstein@masto.deoan.org ("Jeff Sonstein"):
dimsumthinking wrote:

Australia part one- any must dos, must sees, or must eats for Sydney?

I will follow with more cities later

jsonstein@masto.deoan.org ("Jeff Sonstein") wrote:

washing dishes https://www.pandora.com/artist/miles-davis/someday-my-prince-will-come/i-thought-about-you-feat-hank-mobley-wynton-kelly-paul-chambers-and-jimmy-cobb/TRpdzl3zfbm56Zk?part=ug-desktop&corr=8104885

glyph ("Glyph") wrote:

the gell-mann amnesia effect is when you read a journalist writing about some topic that is close to the heart of your own expertise but far from theirs — usually some specific cutting-edge technical expertise — notice that they are incorrect about some details, and then assume that they are also incorrect about a bunch of stuff that they know much better than you do, like political corruption or investigating the consensus of experts on well-settled science

Boosted by glyph ("Glyph"):
cara.city@bsky.brid.gy ("Caræsten 🇵🇸") wrote:

hadn't seen this yet

(though assuredly it's been done)

RE: https://bsky.app/profile/did:plc:wc33ikadg53lobua2ivtxbcx/post/3msgtdqcvnt2s

dysfun@treehouse.systems ("gaytabase") wrote:

relaxed rubiks-balanced tree

Boosted by neatnik@social.lol ("Neatnik :prami:"):
kris@social.lol ("Kris") wrote:

I realized why AI bothers me so much. It takes away the reason to celebrate new achievements of individuals.

Boosted by andrewnez ("Andrew Nesbitt"):
sunnydeveloper ("Emma Irwin") wrote:

Slides from our #FOSSY Talk:

"AI depends on #OpenSource to function, both as a technical dependency and as a source of data. Model builders who consume the output of our labour MUST also be aligned with the values, norms and governance of those communities. Our hypothesis, being that AI builders who prioritize alignment with open source communities will have an advantage built through trust."

https://sunnydeveloper.com/ai-alignment-metrics-for-open-source-communities/

Gargron ("Eugen Rochko") wrote:

I love a good trail.

📷️ Pentax KX
🎞️ Fuji Superia X-tra 400
🔭 Pentax M 50mm/1.7
⚗️ Come Through Lab

#BelieveInFilm #FilmPhotography #AnalogPhotography #35mm #Cumbria #TheLakes #LakeDistrict

glyph ("Glyph") wrote:

like yeah, if somebody is playing Russian roulette at massive scale with a trillion dollars of shotguns tied to elastic cords and industrial motors whipping them around at speed in a giant warehouse I am not saying that this is _not dangerous_, of course it's extremely fucking dangerous, but it's not some terrifying new unforseen capability that our adversaries can leverage against us, it's just a mind-boggling level of normal irresponsibility with dangerous technology that needs to be stopped

glyph ("Glyph") wrote:

the huggingface hack has escaped containment (by which I mean it's getting covered by the mainstream press, not the thing where it literally escaped a container), and it's *so* frustrating how they are all falling for it.

this is a shotgun trigger tied to a jumprope, not an atomic weapon

soatok@furry.engineer ("Soatok Dreamseeker 🔜 DEFCON 34") wrote:

#defcon34 Suiting up one last time!

glyph ("Glyph") wrote:

this product would solve exactly zero problems for me and it would create a whole new category of management headache as well as ongoing costs but I want it so baaad https://amzn.to/4wOWoEj

Boosted by dysfun@treehouse.systems ("gaytabase"):
david_chisnall@infosec.exchange ("David Chisnall (*Now with 50% more sarcasm!*)") wrote:

This story should terrify any open source project or company that allows LLM-generated code into their project.

Someone created a new app, using Claude. Only it wasn’t a new app, it was clearly plagiarised because it happened that there was an example in the training set that exactly matched the requirements. The similarity was well within the range that courts have previously used to determine a derived work.

When you have this level of similarity, the requirement comes to you to prove that there was no way that the original work could have flowed to your project. Companies that have this concern usually do it by ensuring that no one on the team has been exposed to the original and that the code for the original never goes near their systems. But when one of the systems that you use is a language model trained on, among other things, all of the open-source code that it could scrape (and which does not disclose its training set), being able to prove that there was no path from some other codebase to yours is impossible.

Just because the US copyright office has ruled that you, as the person promoting an LLM, cannot assert copyright on the output, does not mean that someone else can’t. If you take a DVD and transcode it to H.264, there is no creative step and so the new copy is not something subject to independent copyright, but it is a derived work of the DVD (itself a lower-quality derived work of the original masters) and so subject to the same copyright.

Importantly in this story, the person prompting Claude had no idea that the original app existed. To safely use the code, they would need to search everything in the training data and discard outputs that would meet the bar of being substantially similar. And that’s something that requires human judgement.

Boosted by glyph ("Glyph"):
whitequark@treehouse.systems ("✧✦Catherine✦✧") wrote:

south korea is having a normal one

Boosted by glyph ("Glyph"):
maxleibman@beige.party ("Max Leibman") wrote:

Weird how nobody's AI has gone rogue, broken containment, and become profitable.