Mastodon Feed: Posts

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
ashur@front-end.social ("Ashur Cabrera 🌺") wrote:

bun team like

Boosted by dysfun@treehouse.systems ("gaytabase"):
nikclayton ("Nik") wrote:

@mitchellh does this mean you're going to remove the "maintainers are exempt" language from https://github.com/ghostty-org/ghostty/blob/main/AI%5FPOLICY.md ?

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
SmudgeTheInsultCat@mas.to ("Smudge The Insult Cat 🐀") wrote:

Boosted by dysfun@treehouse.systems ("gaytabase"):
RosaCtrl@vivaldi.net ("No mother, no bride, no king") wrote:

RE: https://hachyderm.io/@mitchellh/116580433508108130

"its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" means it’s fine to ship bugs because our software most likely shouldn’t exists anyways, it’s absolutely irrelevant, and we are just waiting to cash out

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
olivia@scholar.social ("Olivia Guest · Ολίβια Γκεστ") wrote:

Also a short & snappy paper on these issues, esp. for psychologists and cognitive neuroscientists interested in modelling and AI & in how they contribute to or even hinder research, is this w/ @andrea @Iris

https://doi.org/10.5281/zenodo.19983695

5/

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
olivia@scholar.social ("Olivia Guest · Ολίβια Γκεστ") wrote:

Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia: https://doi.org/10.5281/zenodo.17065099

We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical thinking, expertise, academic freedom, & scientific integrity.
1/n

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

no i'm not vibe codding today, but i have been having thoughts.

most importantly, they make fewer clangers when doing less work at a time. how do you structure everything to make it do less work at a time without having to supervise it constantly? not sure. lots of files?

they are absolutely terrible at iterating. they like to give up and their solution is often to attempt a rewrite rather than fix the errors with what might have been borderline acceptable if they'd just fixed it.

oddly enough this is not an error i have with programming languages - they manage loops just fine. i could for example with the right tools rather easily knock up a loop that builds, runs the tests and otherwise instructs it to continue solving the problem. well the loop bit is easy, at least, but getting the cod to do the thing might be harder.

maybe i can drip feed it errors? it definitely gets confused when receiving a lot of them at a time and what i said earlier about less at a time might work.

so it's not a complete fix, but i think i could probably do a better job if i wrote my own vibe codding apparatus - one that appreciated that there are algorithms other than machine learning available to solve problems. unfortunately this is sounding like a lot of effort to make the silly thing less silly. like an actual engineering problem, in fact.

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

it's sad that the football season is almost over

all we have then is two weeks of more football, followed by a summer of big football, and then finally the next football season

Boosted by db@social.lol ("David Bushell 🪿"):
nicuveo@tech.lgbt ("Antoine Leblanc :transHaskell:") wrote:

i am so tired of "ethical concerns aside" being a phrase i see every single time someone tries to defend the use of LLMs. fuck that! ethical concerns front and fucking center! it is very revealing that tech is currently in such a state that the quiet part can be said out loud without any pushback.

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

A user you’ve blocked has previously contributed to this repository.

Damn, i was just about to try it out

Boosted by dysfun@treehouse.systems ("gaytabase"):
SecureOwl@infosec.exchange ("Mike Sheward") wrote:

BREAKING: GitHub switching status page default language to German in order to return to five neins.

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

think i'll give zig a go again today.

fribbledom ("muesli") wrote:

I've been saying it for years now:

Friends don't let friends buy BambuLab printers 🤷‍♂️

Boosted by dysfun@treehouse.systems ("gaytabase"):
mookie@weredreaming.com ("steve mookie kong") wrote:

Overheard: “I feel more pressure to use AI as an adult than to do drugs as a teenager.”

#ai #llm #drugs

glyph ("Glyph") wrote:

it's that time again #python

Boosted by jwz:
jwz wrote:

@linux I wonder what the monthly cost is for that app that Mike Johnson uses to monitor his porn intake with his son. Maybe he could save some money by rawdogging that, too.

Boosted by jwz:
jwz wrote:

@linux Maybe if Mike Johnson skipped lattes and avocado toast, he could afford to survive on a mere $174,000 per year.

Boosted by jwz:
linux@s.cafe ("Linux and BSD are best 🐧😈") wrote:

Congress House Speaker Mike Johnson says we should have sympathy because Congress isn’t paid enough — so we should “allow” insider trading just so they can take care of their families.

Fuck you, Mike Johnson, and every so-called “public servant” who thinks public service is a get-rich-quick scheme.

#Corruption #Government #Congress

Attachments:

• video: As described in the comment.

Boosted by jwz:
ShadSterling wrote:

@flpvsk @jwz they couldn’t reach a consensus while including ethical concerns, so they decided to exclude ethical concerns and write a policy without considering whether or not that policy is unethical. Their willingness to accept an unethical policy is quite bad

Boosted by jwz:
jalefkowit@vmst.io ("Jason Lefkowitz") wrote:

Someone got paid more money to think this up than I have been paid in my entire life

https://www.theverge.com/news/931918/microsoft-xbox-rebrand-caps

Boosted by aredridel@kolektiva.social ("Mx. Aria Stewart"):
aurynn@cloudisland.nz ("Aurynn Shaw") wrote:

Guilt is not a strategy

Boosted by cstanhope@social.coop ("Your weary 'net denizen"):
ellipticurv3@mathstodon.xyz ("mrityunjai | मृत्युंजय") wrote:

http://existentialcomics.com/comic/652
#Comic #AI

Boosted by jwz:
inthehands@hachyderm.io ("Paul Cantrell") wrote:

I’m kind of speculating here. I get off the LLM coding bus at several earlier stops:

⁃ The energy and water usage are an environmental disaster (so I mostly avoid it for the same reasons I try to reduce my driving).

⁃ The data sourcing is an ethical disaster (so I prefer to avoid it for the same reasons I try to buy fair trade products).

⁃ The people who profit from it at the top are largely horrible (so I’m about as interesting in debating its pros and cons at length as am I debating the work capacity of a Cybertruck).

10/

Boosted by jwz:
atax1a@infosec.exchange ("mx alex tax1a - 2020 (6)") wrote:

every once in a while someone asks us "why can't you just review the slop code as if it were human written" and let me tell you that we have tried it, and you cannot, actually, do this, for the simple reasons that humans do not write code like that

human-written code follows a progression and exhibits a model of the problem domain. the model may be wrong, the progression may be wrong, but line 1 and line 10 are related to each other.

LLM code has none of that. it looks like code, but does not have the structure of it. there is no progression of thought, there is no model of the problem domain, and lines 1 and lines 10 bear no relationship to each other outside of statistical likelihood.

the question thus asks, "why can't you just eat the dogshit as if it were a burger cooked by a human"

Boosted by jwz:
aeva@mastodon.gamedev.place wrote:

i think society would be better if they just handed out money to people no questions asked, but i suppose it's true some people might use it to be happy

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

I just wrote a blog post about recurring subscription billing. (yeah, I know, what an intriguing topic, lol)

https://www.neatnik.net/blog/automatic-renewals/

TL;DR: You can tell me whether I’ve earned your business by actively choosing to renew your membership, not by having to cancel a one-sided assumption that you’d want to renew automatically in the first place.

kornel ("Kornel") wrote:

New Frugit! ( #electroSwing + #cat )

https://www.youtube.com/watch?v=ImS6Yw0O%5FFA

cstanhope@social.coop ("Your weary 'net denizen") wrote:

Whatever, maybe like me you'll enjoy Ei Wada playing a vintage electric fan as a mild distraction from the state of things:

https://www.youtube.com/watch?v=QJavHc48iZU

cstanhope@social.coop ("Your weary 'net denizen") wrote:

All these weird insidious ways of trying to draw you in to something you find immoral.

"If you don't use it, you'll cede the ground to all the people who do use it. Don't ignore it. You can't sway a conversation you're not part of."

Like, my friend, I do *not* have to participate in a morally abhorrent activity to say it is morally abhorrent. If it was just about the conversation, I don't think you'd find it so important to try to make me participate in using or supporting it.

Also, I'm not ignoring it. I am trying to avoid it. You know, like you avoid any bad idea. To avoid it requires you to be aware of it, and to be knowledgeable enough to not be swayed by specious arguments.

Boosted by slightlyoff@toot.cafe ("Alex Russell"):
neurovagrant@masto.deoan.org ("Ian Campbell 🏴") wrote:

@threatresearch @zackwhittaker

THE EXACT MANUFACTURERS AND MODELS. WE'VE KNOWN THEY'D TARGET THESE FOR AT LEAST FIVE YEARS.

FIVE YEARS.

https://news.sky.com/story/irans-secret-cyber-files-on-how-cargo-ships-and-petrol-stations-could-be-attacked-12364871