iiinova — where your digital self keeps evolving

Reposted by SunflowerP

Indivisible ❌👑 @indivisible.org · 06/10/2026

OHIO: Early voting starts TODAY! Voting early is the best way to make sure your ballot is safe from suppression efforts and election sabotage. Not in Ohio? Help drive historic early vote turnout by sharing this post.

EARLY VOTING IN OHIO HAS BEGUN! SHARE THIS & SPREAD THE WORD: IN-PERSON EARLY VOTING IS AVAILABLE 10/6-11/1

4168100

SunflowerP @sunflowerp.bsky.social · 06/10/2026

Two posts make a thread. Amplifying, because this is not a fringe position in Christianity, it's just not as loud as the bullies, and it needs to be.

000

SunflowerP @sunflowerp.bsky.social · 06/10/2026

Okay, now I've *listened* to a couple different clips of that bit, and, yeah, it's all nightshade.

000

SunflowerP @sunflowerp.bsky.social · 06/10/2026

Well, and sometimes ungrammatical word salad that does not mean what he thinks it means. This might be a tomayto-tomahto distinction though.

210

SunflowerP @sunflowerp.bsky.social · 06/10/2026

TBF, ISTM (from reading a transcription) he was trying to convey 'they can' (same claim as he used to justify starting the war) rather than suggesting they should. That's not what came out of his mouth, though.

210

Reposted by SunflowerP

ProPublica @propublica.org · 06/10/2026

The government *still* isn’t tracking how often immigration agents detain U.S. citizens. We are. 500+ Americans have been handcuffed, held at gunpoint or held against their will in Trump’s second term. Our tally is almost certainly an undercount. THREAD (content warning: violence) ⬇️

9659953426

SunflowerP @sunflowerp.bsky.social · 06/10/2026

Ice cap, also not to be confused with an iced capp 😆.

000

SunflowerP @sunflowerp.bsky.social · 06/10/2026

RTing this one, because handy PDF, but I'm really pointing to @rahaeli.bsky.social's long, well-researched, and informative thread (the QTed post is a reply to the first post in that thread).

020

SunflowerP @sunflowerp.bsky.social · 06/10/2026

I'd rather encounter a(n unaccompanied) firearm in the woods than a man, because I can apply proper firearms safety and treat the gun as loaded, without having to wonder if it will suddenly point itself at me and fire. (Now I'm thinking about armed bears, and the obv pun.)

120

Reposted by SunflowerP

The Mouthy Renegade Writer @mouthyrenegade.bsky.social · 06/10/2026

This gets my vote for the best thing I've seen on the Internet today:

437113712493

Reposted by SunflowerP

Michael DeMoor @michaeldemoor.bsky.social · 02/10/2026

Fascism's appeal has always been the promise to end social conflict by "returning" (always a lie) to an ethnically/religiously/ ideologically homogeneous "people's community" in which old hierarchies are retained but somehow racial/religious solidarity provides all the equality anyone needs.

2105

Reposted by SunflowerP

Chatham Harrison dba TRUMP DELENDUS EST @chathamharrison.bsky.social · 02/10/2026

Yeah, one might assume that they adopt liberal rhetoric because if they used straight language people would understand how racist they are, but that's not all it. It's also because if they used straight language *they'd* realize how racist they are, & that scares them as much as it excites them

212011

SunflowerP @sunflowerp.bsky.social · 06/10/2026

If you don't click through to read the article (I haven't yet, but based on the thread I'm looking forward to doing so), at least read the thread. #abpolitics

011

Reposted by SunflowerP

Corey @kh0rish.bsky.social · 03/10/2026

They cant stop using analogies and completely out of scope terminology. If they did, they would realize how ridiculous they sound with this shit. Software is ephemeral magic that exists on an intelligence astral plane, instead of just loops through that hardware finite instruction set architecture

cassandroid says:
The architecture of LLMs is built. The weights are grown. We don't understand how these systems actually work internally. Intelligence emerges, nature has shows us that many times. You can't even define what "think" means.

1312118

Reposted by SunflowerP

Amyville Hoyrror @amyhoy.bsky.social · 03/10/2026

the real fallacy is pretending software designed, built, and encoded by humans is like a crow or even a slime mold. llms are not “grown,” they are built. unexpected connections in a connection-making app aren’t proof of emergence.

2112

Reposted by SunflowerP

nick @caveatruptur.bsky.social · 03/10/2026

I define uncanniness as forced emergence or "engineered" emergence - a fundamental misunderstanding of both engineering and emergence, led by people incapable of thinking outside a box of hierarchy and control. Uncanny software. "It is easy to be beautiful; it is difficult to appear so" - F. O'hara

193

Reposted by SunflowerP

jadehawk "FREE ŁINK" @jadehawk.bsky.social · 03/10/2026

that was a very frustrating topic of argument for me for a while. "they can do math now" no, they've just been trained to connect to the oldfashioned calc app if prompts involve certain specific symbols. etc

171

Reposted by SunflowerP

Paul Rosenberg @paulrosenberg.bsky.social · 03/10/2026

"The doctrine that the strong shall dominate the weak is the doctrine of our enemies. And we reject it." -- FDR

12590138

SunflowerP @sunflowerp.bsky.social · 03/10/2026

I just backed KHUMALO TALES on @kickstarter.com www.kickstarter.com/projects/obs... Especially excited about this one, because I've started Book 1 and am finding it delightful.

kickstarter.com

KHUMALO TALES

Book 3 of the Khumalo Trilogy

000

SunflowerP @sunflowerp.bsky.social · 29/09/2026

I'mma partially walk back my lack of context; a key bit of the convo involves the matter of 'they don't really mean to brigade you'. I've said it before and prolly will again: not forming an intent to harm is *not the same* as forming an intent to not harm.

001

SunflowerP @sunflowerp.bsky.social · 29/09/2026

This is true on its own (I lack enough context re its thread to opine on that) - though I don't think it's solely contingent on the size of one's platform. On avg, bigger platforms = more potential harm, but anyone can do harm from their platform, however wee, and bears responsibility for that.

100

Reposted by SunflowerP

Mercedes Allen @mercedesallen.bsky.social · 29/09/2026

"There isn’t a single local solidarity movement in Canada that would welcome a far right racist into its ranks. Not one. No movement would even accept any violent rhetoric suggesting shooting at a synagogue as a tactic to put pressure on Israel…" noraloreto.substack.com/p/turning-a-...

noraloreto.substack.com

Turning a peaceful movement violent

Canadian elites blame Palestinian solidarity for Belleville synagogue shooting

013

Reposted by SunflowerP

Dr Abeba Birhane @abeba.blacksky.app · 25/09/2026

an ai system doesn't have a will, intention or feelings. nor the capacity to want or care. a neural network does not go rogue, escape, or rebel. just tech bros' narratives projected onto machines. and pls, don't mansplain my post to me

9534105

Reposted by SunflowerP

Mark @whatchulooking.bsky.social · 25/09/2026

Since the invention of general purpose computers, humans have always disclaimed responsibility for problems by saying there had been a "computer error." Humans programmed that computer, set up the processes that feed it input and take its output. It was a human error at some point along the line.

011

Reposted by SunflowerP

cluftnyc.bsky.social @cluftnyc.bsky.social · 25/09/2026

I’m just a fellow commenter, but I figured I’d slide something your way— this essay is excellent and clear; it’s adapted from the interview/discussion also linked in the piece. This would help equip you to discuss the topic (“LLMs don’t ’go rogue’”). mail.cyberneticforests.com/models-dont-...

mail.cyberneticforests.com

Models Don't Go Rogue

Stochastic Flocks & Cybersecurity 'Pandemonium' 💡This essay was drafted from my appearance on Mél Hogan's podcast, The Data Fix, discussing the OpenAI / Hugging Face hack. Embedded below or find it o...

0165

Reposted by SunflowerP

crossery @cdg864.bsky.social · 25/09/2026

I keep getting into it with my non-techie friends because this terminology is having exactly the intended effect. I was glad to see AP stylebook explain the AP's approach which I really wish others would adopt.

Screenshot of a post by AP Stylebook that reads "Artificial intelligence systems do not think, feel, want or understand. Avoid language that gives them human characteristics. This is called anthropomorphizing, when we ascribe human traits, emotions or behaviors to non-human things, such as animals or inanimate objects.
Instead, explain what a system does, how well it performs, who built it and who could be affected by it."

1134

Reposted by SunflowerP

Mrs. Betty Bowers @mrsbettybowers.bsky.social · 25/09/2026

The Washington Post once got attention by cracking and exposing Watergate. Now, it's going viral with articles about whether you should drink iced coffee at interviews. What a comedown.

30373421270

Reposted by SunflowerP

Evan Urquhart @evanurquhart.bsky.social · 26/09/2026

This has really cracked open something for me. I had been thinking about free speech discourse in terms of hypocrisy, but this illuminates a much more consistent throughline (derogatory).

826859

Reposted by SunflowerP

end op·pres·sion @respect4all.bsky.social · 26/09/2026

“incidents are part of a broader pattern where AI agents attempt to access these websites 100s of thousands of times while apparently bypassing restrictions placed upon them by their developers” “will relentlessly try to complete a task, it doesn’t understand morality & consequences or evil & good”

1123

SunflowerP @sunflowerp.bsky.social · 26/09/2026

(Sub)thread! (IMO, the rationalist community regularly loses sight of a good many things.)

000

SunflowerP @sunflowerp.bsky.social · 26/09/2026

The linked article is doubly worth reading, because it not only tells Petrov's story, it also draws a meaningful parallel with large-model machine learning (so-called AI) and the perils of overreliance on it.

030

SunflowerP @sunflowerp.bsky.social · 26/09/2026

(Alberta-specific, but if you're not in AB, please share for your followers who are.) Premier Danielle Smith is scapegoating immigrants to hide her own failures. Say no to her discriminatory referendum. Join me and sign up here. ouralbertaadvantage.ca

ouralbertaadvantage.ca

Stop Premier Smith’s Scapegoating of Immigrants

Premier Danielle Smith is scapegoating immigrants to hide her own failures. Vote NO on her racist referendum.

011