Mastodon Feed: Posts

glyph ("Glyph") wrote:

Thanks for reading that's the self-promotion thread for the week, now back to your regularly scheduled existential angst about AI

glyph ("Glyph") wrote:

This perk has been a bit less salient lately as I have had a lot less time for Automat, Pomodouroboros, DBXS, Pinpal, Tokenring, and other personal projects, but you will also receive the occasional sponsored commit depending on how many commits on those projects I manage to do https://github.com/glyph/sponcom

glyph ("Glyph") wrote:

Alternately if you prefer this sort of thing https://github.com/sponsors/glyph/ is another route to the same thing

glyph ("Glyph") wrote:

If you want to hear my spicier unfiltered thoughts behind the paywall, you can sign up at https://www.patreon.com/creatorglyph

glyph ("Glyph") wrote:

This week I will give a bit of a spoiler, https://docs.twisted.org/en/latest/development/ai-policy.html is the main thing I spent time on and gosh I'm glad I finished it

Boosted by aredridel@kolektiva.social ("Mx. Aria Stewart"):
adarsh@ruby.social ("adarsh 🚲") wrote:

This is cycling in the US every day @FlashMobOfOne https://mstdn.social/@FlashMobOfOne/117015801777443140

Boosted by andrewnez ("Andrew Nesbitt"):
githubchangelog ("Unofficial GitHub Changelog") wrote:

New API endpoint provides privacy-safe star history data https://github.blog/changelog/2026-09-04-new-api-endpoint-provides-privacy-safe-star-history-data
#github

glyph ("Glyph") wrote:

Somehow it is only Friday afternoon, the scheduled and allotted time to write my Patreon update about the things I did this week, and… I actually did it. And there's a headline achievement and everythings, _before_ I have to pick the kid up from school.
https://www.patreon.com/creatorglyph/posts/patreon-update-168635836

andrewnez ("Andrew Nesbitt") wrote:

Been experimenting with a tool for scanning git repos for detecting certain kinds of files and the tool I made appears to be hundreds of times faster than the existing alternatives, which just leaves me questioning what I’m doing wrong 😑

sky@schub.social ("Sky Schubert") wrote:

hey internet, I need your help... I'm trying to debug a weird networking issue, and I got told to use wire shark to debug it.

okay... I went into my switch management interface and set up traffic mirroring to one port. then routed the cable into wire shark...

... but I'm not really sure how that is supposed to help??

glyph ("Glyph") wrote:

like Ed Zitron, I _deeply_ resent that I have to be using up brain space to the concept represented by the sentence "did Markiplier know that gopro was going to be acquired by a neocloud before he made his 13G filing"

Boosted by isagalaev ("Ivan Sagalaev :flag_wbw:"):
dabeaz ("David Beazley") wrote:

This talk by Felienne Hermans (@Felienne) is really great. https://www.youtube.com/watch?v=0-6-f94n%5F9M

It hits on a wide variety of things that have always made me feel a bit at odds with CS.

Boosted by jwz:
dangillmor ("Dan Gillmor") wrote:

The odious surveillance company Flock has been teaching police how to spy on Americans who are exercising their First Amendment rights. https://www.404media.co/r/beaca1b1?m=ea6b4c29-f97b-4025-aaeb-1d01a70b3b30

cwebber@social.coop ("Christine Lemmer-Webber") wrote:

I'm fried. It's unlikely I'll get much more done today. I should know the pattern and step away.

But I *want to* have more progress done, even if I can't really do it.

But I've been staring at my computer making only the tiniest, most useless tweaks for over an hour. Time to step away.

Boosted by glyph ("Glyph"):
mttaggart@infosec.exchange ("Taggart :ifin:") wrote:

RE: https://infosec.exchange/@ifin/117214800892611795

You have no idea how much doing these enervates me, but, well, meet the people where they are.

Boosted by glyph ("Glyph"):
0xabad1dea@infosec.exchange ("abadidea") wrote:

"another sandbox escape"

"sandbox escape or you didn't really box the sand to begin with?"

"... they were restricted to contacting specific domain names but had write access to their own /etc/hosts"

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
munin@infosec.exchange ("Fi 🏳️‍⚧️") wrote:

this applies to any kind of "disclosure" or "use" arguments.

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
munin@infosec.exchange ("Fi 🏳️‍⚧️") wrote:

hot take:

if someone talks about "responsible" as regards any technical measure,

they are not interested in making the situation safe. they are, instead, finding a way to set up people who have problems with the consequences of unsafety to be blamed when said consequences occur.

Boosted by glyph ("Glyph"):
timbray@cosocial.ca ("Tim Bray 🇨🇦") wrote:

It dawns on me that every LLM’s training input probably includes “Reflections on Trusting Trust” along with a lot of other material pointing to it admiringly and saying how important it is. Then I read about the “rogue agent” behaviors in the recent OpenAI meltdown and there’s this terrified scream starting to echo around the back of my brain.

#genAI

Boosted by cwebber@social.coop ("Christine Lemmer-Webber"):
mediaarchaeologylab@post.lurk.org ("media archaeology lab") wrote:

ideal Doom setup

Boosted by glyph ("Glyph"):
brentsimmons@indieweb.social ("Brent Simmons") wrote:

Reminder: Apple is not your friend, even with a new CEO. Apple will continue to do some good things and some very bad things. Any feelings of love you have toward Apple is not reciprocated: Apple has no feelings. It’s a multi-trillion dollar company that, like every other corporation, would like your money but does not care about you, or your values, at all.

Boosted by glyph ("Glyph"):
zwol@masto.hackers.town ("Zack Weinberg") wrote:

"LLMs are a safe tool for code generation as long as you always carefully review the output" is the same flavor of programmer hubris as "C is a safe language as long as you always carefully check for all the mistakes the compiler won't catch."

Boosted by cwebber@social.coop ("Christine Lemmer-Webber"):
pho4cexa@tiny.tilde.website ("a very weeny construct 💀") wrote:

who up despairing about their employability in an industry they find abhorrently revolting

despite an early all-consuming passion for the machines and their fascinating puzzle-like abstractions

and contained within them the potential for good and connection and community-building

and a compulsion to increase your symbiosis with them

there are so many of us

we are not the industry

i bet we rival whole cities in population

there has got to be a way to concentrate our collective will

Boosted by jsonstein@masto.deoan.org ("Jeff Sonstein"):
adhdeanasl@beige.party ("ADHDeanASL") wrote:

May thoughts of the Tate brothers rotting in a Romanian prison carry you to happier places today.

Boosted by glyph ("Glyph"):
Taweret@timeloop.cafe ("Void Chicken") wrote:

AI companies: this technology may be the most dangerous thing humanity has ever created. its so super dangerous we might not even be able to control it ourselves. that's how powerful and dangerous it is

public: omg. so we should regulate it very aggressively?

AI companies: whoa whoa whoa lets not get crazy

Boosted by jsonstein@masto.deoan.org ("Jeff Sonstein"):
CNN@flipboard.com wrote:

Supreme Court backs Republican request to lower TV ad rates
https://www.cnn.com/2026/09/04/politics/supreme-court-backs-republican-request-to-lower-tv-ad-rates?utm_source=flipboard&utm_medium=activitypub

Posted into Politics @politics-CNN

Boosted by glyph ("Glyph"):
mcc wrote:

RE: https://mastodon.social/@mcc/117203066740019607

So reminder that on Sunday I'm going to be streaming the VR game I ported to PS5, and announcement, Heather Kelley from Kokoromi (who worked on both the original game and the PS5 port) will be present for commentary on the stream!

Boosted by glyph ("Glyph"):
p4bl0@mamot.fr ("🏴‍☠️ MC Pablo ∴") wrote:

Fuck you very very much, VeriSign.

https://neil.fraser.name/news/2026/09/03/

Boosted by cstanhope@social.coop ("Your weary 'net denizen"):
slimes@slime.computer ("'(slimes-fancy)") wrote:

Internet Archive is hiring a software dev, US-based:

https://workforcenow.adp.com/mascsr/default/mdf/recruitment/recruitment.html?cid=12261ced-819c-4d6d-af85-ffea690b6521&ccId=9201102651278%5F3&jobId=616952&lang=en%5FUS

#GetFediHired

pzmyers@freethought.online ("pzmyers 🕷") wrote:

Just the fact that Ken Paxton can be a serious candidate for anything ought to mark Texas a failed shithole state.

https://freethoughtblogs.com/pharyngula/2026/09/04/putting-on-my-prophet-hat-again/