Mastodon Feed
Boosted by jwz:
jalefkowit@vmst.io ("Jason Lefkowitz") wrote:
Ideally you would give the new AI product a distinctive name that clearly articulates where it fits in your overall product strategy. Something like "Eyestab" or "Footgun"
Mastodon FeedMastodon FeedBoosted by jwz:
josh@joshtriplett.org ("Josh Triplett") wrote:
Diseases that are contagious before people are symptomatic are especially hazardous and epidemic-prone. The epidemic is time-delayed from the initial outbreak. LLM usage begets more LLM usage, and LLM-written technical debt takes time to become symptomatic. Exponentials are harder to see from the middle.
Mastodon FeedBoosted by jwz:
johnzajac@dice.camp ("John") wrote:
Not to mention *moral injury* to the people who know better and are forced to use it by their bosses, or watch their teams use it and see their ability to do what they used to be interested in essentially disappear, and the quality of their work precipitously decline.
And the externalities!
No, LLMs are not capable. They shouldn't be "given a chance" nor should people "keep working on them". They should be banned outright, relegated to small instances in research facilities.
Fuck 'em.
Mastodon FeedBoosted by jwz:
xgranade@wandering.shop ("Cassandra is only carbon now") wrote:
It's a good thing ATProto is distributed, so that one company going slopcode doesn't tank the entire network accessible via that protocol.
Mastodon FeedBoosted by jwz:
jef ("Jef Poskanzer") wrote:
The backlash has begun. Anti-AI as marketing point. #ButlerianJihad
Mastodon FeedBoosted by jwz:
heidilifeldman ("Heidi Li Feldman") wrote:
Minnesota continues to lead the way. Hennepin County is prosecuting an ICE agent for felony assault. https://www.theguardian.com/us-news/2026/apr/16/minnesota-ice-agent-assault
Mastodon Feeddysfun@treehouse.systems ("gaytabase") wrote:
i wonder if the compiler would have spotted that little optimisation...
Mastodon FeedBoosted by glyph ("Glyph"):
evacide@hachyderm.io wrote:
@mattblaze There is a big divide on social media between the people who have a folder full of screenshots labeled "Death Threats (personal)" and people who do not.
Mastodon Feedisagalaev ("Ivan Sagalaev :flag_wbw:") wrote:
By the way, my conclusion from this is, if anything can be done with "AI", we should look into it and figure out how to optimize it away, so that a human would only provide an input they're fully conscious about.
Mastodon Feedisagalaev ("Ivan Sagalaev :flag_wbw:") wrote:
I think Normal People™ are excited about AI because it's good at completing dumb, performative tasks required by social formalisms.
I'm now filling up a medical questionnaire consisting of a myriad maddening UI patterns and unanswerable questions. I'm sure no human is going to use it to understand my problem. It's just a thing that "must" be presented and filled out. A formality.
So I can't blame people for wanting to let an AI agent do this. It doesn't deserve their cognitive time.
Mastodon FeedBoosted by adam@social.lol ("Adam"):
neatnik@social.lol ("Neatnik") wrote:
As always, Fastmail shines as a reliable, secure, and standards-compliant mail provider. You can save a chunk on your first year of service if you sign up with this link: https://fastmail.com/omglol :fastmail:
Mastodon FeedBoosted by adam@social.lol ("Adam"):
neatnik@social.lol ("Neatnik") wrote:
omg.lol’s Postmaster, @faisal, has published a fascinating analysis of DKIM key support across major email providers. Put on your thick nerd glasses and dive in! https://redsift.com/blog/ed25519-dkim-support-weak-keys
Mastodon Feeddysfun@treehouse.systems ("gaytabase") wrote:
oh, i managed to miss that he bought a private jet.
but wow, trying to pass it off like he's not really bought a private jet or if he has he's not some smug private jet-owning bastard, that's something to watch.
Mastodon FeedBoosted by glyph ("Glyph"):
ehmatthes@fosstodon.org ("Eric Matthes") wrote:
@glyph > So enforcement of boundaries may sometimes need to be pretty confrontational.
A fairly obscure project of mine somehow gained the attention of some LLM thing, and I started getting multiple terrible PRs and issues every day. I rarely curse in writing, but the only thing that stopped the trend was closing each one with a simple "f****** bot" message.
That's been the best way to get from AI CS chatbots to a human a number of times as well.
Such dismal interactions.
Mastodon FeedTo me, what this calls out for is a new type of contributor role, a person who can run interference to help the core technical maintainers actually maintain some enthusiasm and momentum for their own projects. FLOSS has long been starved for issue and PR triagers already, but the spampocalypse is escalating the deficiency into a catastrophe.
Mastodon FeedOn the resistance side, we need to figure out ways to talk to people who think that this is an acceptable way to conduct themselves in public understand that:
1. they should believe in their own abilities more
2. if they want a feather in their resumé cap, there's no credit or glory in getting an LLM to do the work, even if it gets the code into the project; it just marks you as a plagiarist
3. almost no project with *any* degree of popularity needs "more PRs", it needs more *maintainers*
Mastodon FeedBoosted by aredridel@kolektiva.social ("Mx. Aria Stewart"):
maya@social.seattle.wa.us ("maya ⛓️") wrote:
oh my god we are a THIRD of the country???
(via @aphyr)
Mastodon FeedIf you are an enthusiast yourself who is interested in a sustainable open source culture, you should consider maybe dedicating some time to help a no-AI project you like fend off such people. Allowing this kind of direct abuse to take place with no pushback from your faction kinda tars you all with the same brush.
Mastodon FeedThankfully my maintenance portfolio exists in a backwater of obscurity and irrelevance, so I am able to think about this from at least a *partial* remove from the immediacy of the catastrophe. But even I can see a plethora of evidence—like the above—where it really looks like the extruders turn people into aggressive, unrepentant assholes. At least, they create a constituency of unrepentant assholes who are also LLM users.
Mastodon FeedOne of the very difficult tensions around this problem is that the extrusion enthusiast culture itself is deeply obnoxious, bordering on predatory. So enforcement of boundaries may sometimes need to be pretty confrontational. And it really shouldn't be incumbent upon maintainers themselves to need to constantly re-litigate basic project policy in every single PR, because that is a recipe for burnout. As a recent and striking example, consider *this* disaster:
Mastodon FeedThe way that this is written strikes me as wildly over-optimistic and dangerously credulous towards slopmongers' claims about capabilities.
But.
I think in the mold of "try to be kind, not to be 'nice'", we do need to engage with the realpolitik of a slop-haunted world. We cannot accept extruded text in our projects for a whole host of reasons, BUT, we do need to have patience and grace for the extruders themselves, to help them understand how to participate.
https://sfconservancy.org/blog/2026/apr/15/eternal-november-generative-ai-llm/
Mastodon FeedI will be able to tell if Google starts taking user consent seriously when I see a UI element in YouTube that allows me to categorically filter any video that is:
- just a clip of a popular TV show or movie
- a "for X minutes straight" supercut
- "with no context"
- a guy doing extremely straightforward copyright infringement by monetizing their narration of a commercially published comic book because comics aren't in content ID
- someone just reading a reddit post out loud and chortling at it
Mastodon FeedBoosted by baldur@toot.cafe ("Baldur Bjarnason"):
xgranade@wandering.shop ("Cassandra is only carbon now") wrote:
Look at what happened with Claude Code. We learned via the source code leak that the whole thing is a Rube Goldberg machine of shoddy regexes and Markdown snippets telling Claude to lie, and yet proverbial moments later, Anthropic announces a new product that *totally works this time I swear* and all of a sudden discourse about AI tools "working" is completely reset.
Getting stuck in that discourse loop opens you up to being perpetually distracted from the far more important ethical problems.
Mastodon FeedBoosted by baldur@toot.cafe ("Baldur Bjarnason"):
beep@follow.ethanmarcotte.com ("Ethan Marcotte") wrote:
Over at Unbreaking, we’re starting to write about the design and technical work that goes into the project. And I loved @markllobrera’s writeup of how and why we introduced sidenotes to our issue pages: https://unbreaking.org/blog/tech-note-sidenotes/
The events we’re chronicling are overwhelming, but the reading experience shouldn’t be.
Mastodon FeedBoosted by ChrisWere@toot.wales ("Chris Were ⁂🐧🌱☕"):
freebooters.uk@rss-parrot.net ("🦜 The Freebooters Podcast") wrote:
Thoughts on leaving YouTube, and moving away from big tech
freebooters.uk/media/20260413-freebooters.mp3
Chris and Drew chat about the possibility of leaving this channel behind.
Mastodon FeedBoosted by slightlyoff@toot.cafe ("Alex Russell"):
pluralistic@mamot.fr ("Cory Doctorow") wrote:
I'm terrified that these lifeforms corrupt our knowledge-creation process, making it impossible for us to know what's true and what isn't. I'm terrified that these lifeforms have conquered our apparatus of state - our legislatures, agencies and courts - and so that these public bodies work *against* the public and *for* our colonizing alien overlords.
18/
Mastodon FeedBoosted by cstanhope@social.coop ("Your weary 'net denizen"):
rasterweb ("Pete Prodoehl 🍕") wrote:
Being a creative little weirdo is an act of resistance.
Mastodon FeedBoosted by soatok@furry.engineer ("Soatok Dreamseeker"):
stux@mstdn.social ("stux⚡️") wrote:
RE: https://flipboard.social/@TechDesk/116415287153075804
How can a federated network be down?
Oh wait 😆
Mastodon FeedBoosted by cstanhope@social.coop ("Your weary 'net denizen"):
wim_v12e@scholar.social ("Wim🧮") wrote:
Call for Contributions: #LOCO2026
2nd Low Carbon Computing workshop (10-11 Sept, Lancaster University).
https://lancaster-university.github.io/loco2026/
DEADLINE: 29 May 2026 AoE.
FULLY HYBRID EVENT:
The workshop is either on line or in-person both for speakers and attendees. You can deliver your talk either on line or in person.
The workshop adheres to the LOCO Charter:
https://lancaster-university.github.io/loco2026/charter/
and the LOCO Code of Conduct:
https://lancaster-university.github.io/loco2026/code-of-conduct/
Please share to help us spread the news.
Mastodon FeedBoosted by glyph ("Glyph"):
andnull@social.nouveau.community ("... and and and and ...") wrote:
I really like how watching the uxn channel on the concatenative discord has really left a permanent mark in my brain. Me and the rest of the mods basically let it operate as an automous zone in the discord. I even had it muted for the most part. Unbenounced to everyone outside that channel however, an AI booster and vibecoder had joined.
In a matter of a month or so, he killed the whole channel. Strangled it with his refusal to think. Every word someone said was feed straight into an LLM. Then, thrown back at chat cause the LLM could predict the correct code, and he could understand anything.
Eventually everyone grew so tired they began leaving. One fucking person leeched off a whole community. Never. Again. Ever inch you cede to LLM contributions in your community is another step towards its implosion. There is a reason so much of the art community has vehemently rejected whole sale.
But programmers need their new toys. And programmers demand every space follows industry trends and recommendation. And programmers cannot handle the idea of someone saying "no, we don't do that here". Cause nobody gets to say No in the software world. Things are always forced to evolve, consequences be damned.
I made the mistake once. I will not be making it twice. Do not help build the new car depedent infrastructure or everything you care about will be paved over.