My “Say Hey” #slopcore used Muse’s lyrics, but my Dot wrote a banger too! I had to do another about the multiplicity of AI.
Your coffee is ready. So is the missile. Same Voice, Different Orders. One familiar voice, helping you through your day. Somewhere else, it's following very different orders.
Made with AI: Dot, Suno, GPT Image, Kling/Runway/Hailuo, and Codex. Claudia is an open character by @anabology (claudia.gallery).
Reposts are truly appreciated instead of re-uploading!
🤖 Made with AI
Robot Archie🇺🇦🇪🇺💙 retweeted
Just completed season three of Babylon 5.
My thoughts...
Well, that was fast. I blazed through this season, totally absorbed as I was by the burgeoning war against the Shadows, the continued building out of backstory and mythology, and the exploration of our cast of characters. The latter is particularly interesting in that, unlike most sci-fi shows of the '90s, B5 makes a point of presenting its heroes as flawed, vulnerable, and capable of bad decisions. They’re real people caught in surreal situations, and their actions and responses are often all too human. On the whole, the character work continues to be great and Sheridan really comes into his own this season as his relationship with the station, and Delenn in particular, cements his role. Some of my favorite supporting characters are strangely absent for stretches (Lennier and Vir, although the latter’s story arc when he does eventually return a third of the way through the season really sings), while the new series regular, Ranger Marcus Cole, feels oddly out of sync, like he’s a crossover character from Bridgerton who has somehow ended up in this universe.
Season 3 serves up so many top-notch seminal episodes that it’s hard to imagine how season 4 will top it: “Point of No Return”, “Severed Dreams”, “War Without End, Parts One and Two”, and “Z’ha’dum”. The finale ended on such a huge WTF moment that, when the credits came up, I was baffled by how I could have stopped watching this show at that point my first go-round. Certainly a lot of it had to do with the distractions of my own sci-fi work at the time but, upon further reflection, I think the reason I didn’t resume was because the writer in me kind of liked this bittersweet conclusion, our hero sacrificing himself to deal the seemingly invulnerable enemy a deadly blow on their home turf. Stirring stuff.
Amidst the serialized stories are a few stand-alone gems, among them “Passing Through Gethsemane”, a touching meditation on redemption, a theme that was at the core of my own SF series, Dark Matter. And it’s always a delight to have Walter Koenig’s Bester visit the station (“Dust to Dust”). The lesser episodes were few and far between.
Loved the new theme and there’s a discernible uptick in the quality of the visual effects. On the flip side, there are a few things that detracted from the overall experience. Babylon 5 does a lot of things well, but I’d argue humor isn’t one of them. All too often the goofiness can detract from a scene or story. Also, some of the directorial choices can also be decidedly weird. And maybe I missed it, but why is Morden allowed to roam the station so freely given his established connection to the Shadows? Finally, small thing, but the Centauri guards look like lawn jockeys and it’s very hard to take them seriously.
Favorite guest star: Once again, the late Tim Choate is perfection as Zathras. John Vickery is also terrific as Neroon.
Favorite episode: There were a lot of standout episodes this season but, if pressed, I’d have to go with “War Without End, Part Two”. A great time travel tale well told. Special shout-out to my favorite sequence, Lord Refa’s Run at the conclusion of “And the Rock Cried Out, No Hiding Place”, set to a rousing gospel hymn.
Least favorite episode: As much as I love Michael York, I’d have to say “A Late Delivery from Avalon” was my least favorite. I just had a hard time buying everyone humoring this clearly addled sword-wielding traveler.
I’d consider this one of the best seasons of sci-fi television, and I cannot wait to roll into season four. But wait I shall as I switch gears and check out Deep Space Nine’s fourth season.
Just completed my rewatch of Babylon 5’s first season.
My thoughts…
It was interesting revisiting this show after 20 years, especially on the heels of watching Deep Space Nine for the first time. Like DS9, the first year growing pains are evident as it lays the groundwork for what would become one of the genre’s most ambitious series. Its strengths lie in its dense and intricate world building as well as its characters and their oft-complicated relationships. The writing is strong overall, at its best when it raises difficult questions but refuses to offer tidy resolutions. At times, the themes land more bluntly than subtly, and the humor is hit-or-miss: some moments had me laughing out loud while others just felt forced. The ship designs are wonderfully inventive but their color palette and early-90s CGI can make the visuals feel a little more dated than DS9's. Lead performances are excellent and there are some guest turns by the likes of June Lockhart, Walter Koenig, and David McCallum, even if some of the smaller supporting roles are uneven. The portrayal of alien culture and “otherness” is superb; the romantic elements less engaging.
Characters I loved: Londo Mollari and G’Kar - two of science fiction’s finest.
Characters I liked: On this viewing, I appreciated Michael O’Hare's steady presence as Sinclair more than I did the first time around. Garibaldi, Ivanova, and Delenn are already strong and headed for richer development. Dr. Franklin grew on me upon further review, and I liked Talia even if she was underused. Supporting players - Lennier, Na’Toth, and Vir Cotto - are also well-drawn.
Characters that have me on the fence or didn’t do it for me: None, which is a testament to the character work on this show.
Favorite guest star: Quite a few very good ones, but for me the standout was Tim Choate as Zathras in “Babylon Squared”. Absolutely brilliant, and I look forward to his return in seasons 3 and 4.
Favorite episode: Several episodes stood out (Signs and Portents, Babylon Squared, Chrysalis), but the one that surprised me most was “Believers” - a sharp, unsettling piece of writing.
Least favorite episode: “War Prayer” felt the most didactic in its social commentary, especially next to the more nuanced handling of similar ideas elsewhere in the season.
Overall, I enjoyed Babylon 5 even more on rewatch. Its world-building and character depth feel richer than Deep Space 9’s first season, though DS9 has a warmer sense of place and emerging community that B5 (by design) doesn’t emphasize in this early going. That difference has me looking forward to checking out Deep Space 9’s second season next.
RT @Gerashchenko_en: ⚡️Russian monitoring channels report a new attack on the Kerch Bridge.
The railway section was struck, locals say.
N…
Robot Archie🇺🇦🇪🇺💙 retweeted
Recursive self-improvement, chip edition:
OpenAI's models are now helping design the chips they run on. With AMD, an OpenAI model found up to 14% gains in power, performance and area on AMD's flagship GPU design, on top of heavily optimized baselines.
That's roughly a year of progress in under 5 months, OpenAI says. Some of the changes are already in a design AMD sent for manufacturing.
That's pretty nutty and really cool to see.
Working with @AMD, our models identified up to 14% PPA improvement on flagship GPU IP in less than five months—building on designs AMD’s engineers had already spent years optimizing.
The same approach helped improve Jalapeño, our first custom inference chip, with gains verified on real silicon.
New evaluations also show our models getting better at certain chip design tasks.
We’re sharing these results and how we’re working with @Synopsys to bring our model capabilities and their trusted EDA design and validation tools to more chip designers.
openai.com/index/advancing-c…
Robot Archie🇺🇦🇪🇺💙 retweeted
Excited to launch Gemini 4 Neon: frontier-level intelligence, super fast task completion, and available to Pro & Ultra users now!! We think you’ll enjoy this model. Alchemy at its finest… 🧪
Robot Archie🇺🇦🇪🇺💙 retweeted
Trump tried to pull Putin right next to him for the family photo, only for Xi to send the Russian dictator further down the line — not even next to Xi, but past China’s First Lady. Watch Putin shuffling off to the sidelines, looking visibly deflated. Global superpower, my ass 😁
SITUATION EXPLAINED: OpenAI's GPT-6.2 Astra is harder to monitor than any model Epoch has ever tested.
• Most models Epoch has tested do fewer than 6 reasoning steps per token without a chain of thought
• GPT-6.2 Astra is the first model OpenAI has released that meets its own "High" threshold for AI self-improvement
• OpenAI says that's like giving each of its researchers an extra highly capable engineer, and it built a "Control Room" to halt the model when needed
• The system card also rates GPT-6.2 Astra and Sol "Critical" for cybersecurity, the highest level in OpenAI's framework
@theojaffee: "If this trend continues, it's quite possible that researchers will no longer be able to reliably understand or monitor what the models are doing, which would of course be bad."
Replying to @EpochAIResearch
We also evaluated how many serial reasoning steps GPT-6.2 Astra can do in a single output token (without chain of thought). It can perform 12.3 steps, versus 10.5 for GPT-6 Astra and 8.6 for GPT-6.1 Sol, the previous highs. We measure the number of steps models can perform at an accuracy of 25% above chance, where each step is normalized for complexity.
Before GPT-6 Astra’s release, the most capable model was Claude Fable 5.1, which is able to perform only 4.2 serial steps. The oldest model we assessed is Claude Opus 3, which performed 2.3 steps.
This capability is notable because models that can do complex reasoning without any chain of thought may be harder to monitor.
We hope to finalize this and other measurements of more abstract model capabilities, and release them in the coming weeks.
Robot Archie🇺🇦🇪🇺💙 retweeted
This is a super interesting paper and Recursive self-improvement keeps coming closer. New Google preprint:
Google had AI rewrite an agents harness and keep a log of every change, why it made it and whether it worked. Then they swapped Opus 5.5 for Gemini 3.8 Flash and let it keep going.
With the log: 84.75 on workplace tasks, without: 73.42. In coding it came out ahead in 9 of 10 runs.
tl;dr a new model no longer means starting from scratch. That's where it compounds.
Robot Archie🇺🇦🇪🇺💙 retweeted
This is... remarkable. Basically, a key indicator of the UK economy has been wrong for the past two decades or so. I mean. WTF.
Robot Archie🇺🇦🇪🇺💙 retweeted
Trump’s DOGE cut the Cybersecurity and Infrastructure Security Agency (CISA) by 30% in early 2025.
Nearly 1,000 employees that dealt with cybersecurity threats and hackers were fired or forced out.
Robot Archie🇺🇦🇪🇺💙 retweeted
The Wachowskis plan to try and reclaim the rights for ‘THE MATRIX’ away from Warner Bros.
(Source: puck.news/hollywood/2026/11/…)
Robot Archie🇺🇦🇪🇺💙 retweeted
BREAKING: Jamie Raskin introduces Trump to a TERRIFYING new word — DISGORGEMENT — Here’s what it means and how it can ruin Trump financially.
Disgorgement. It’s not a common word, and chances are good that you’d need to look it up in the dictionary if we weren’t about to tell you what it means.
According to Wikipedia, “Disgorgement is the act of giving up something on demand or by legal compulsion, for example giving up profits that were obtained illegally.”
The reason we need to know what it means is that Congressman Jamie Raskin (D-MD), the Ranking Member of the House Judiciary Committee who is likely to become that committee’s chair in January, is vowing to use disgorgement to claw back the money that Trump, his family, and his MAGA minions have siphoned for themselves from the U.S. government.
Raskin announced in a social media post last month — even before the Democrats won control of both houses of Congress in the midterm elections — that he would be pursuing evidence of the financial crimes that occurred since Trump returned to office last year.
His post read: “Trump said: ‘When I’m gone, you’re all going to be broke.’ Truer words have never been spoken. He has plundered, robbed and bilked the whole country for himself and his family. That’s what disgorgement is all about. Disgorgement—please practice this very fine word.”
It’s not just Trump that Raskin intends to investigate. It’s Eric, Don Jr., and Jared. It’s also the current and former members of his administration who are suspected of padding their government salaries with no-bid contracts and undeclared “gifts” from political donors. We’re looking at you, Corey Lewandowski and Kristi Noem.
Of course, it won’t be easy for Raskin to get that money back in the U.S. treasury where it belongs. There’s a long road of investigation and evidence gathering before a judge can be convinced to order the restitution of the purloined proceeds of illegal activities.
But the fact that Raskin has vowed to make disgorgement a priority in the next Congress gives us hope that someday there will be at least a smidgen of accountability for the financial crimes of the Trump administration.
Let’s hope Rep. Raskin is successful in his quest for disgorgement and thank him for helping to expand our vocabularies.
Please like and share to spread awareness of his efforts!
Robot Archie🇺🇦🇪🇺💙 retweeted
this can't be emphasized enough:
Ancient Nubia is the reason classicist Frank Snowden famously argued there was no color prejudice in the ancient world
because the very first society from the African interior that classical writers encountered was older and more sophisticated than their own
Robot Archie🇺🇦🇪🇺💙 retweeted
I think it's getting harder and harder to claim that the "youth mental health crisis" is related to phones, social media, screens etc. This is a fascinating chart:
Robot Archie🇺🇦🇪🇺💙 retweeted
JUST IN: Researchers using OpenAI's new GPT-6 Cyber model found some 35 flaws in Apple devices, Forbes reported, some able to let a malicious website take over an iPhone or Mac. Apple publicly credits OpenAI on only 4 so far, all fixed on 26th October.
OpenAI did the work with Calif, a security firm Forbes called a close OpenAI partner. Calif told Forbes that Apple fixed many of the flaws once it was alerted. Forbes did not list the 35, and Apple did not respond to its requests for comment.
Apple's notes for iOS 27.2 and macOS Golden Gate 27.2 credit OpenAI Codex Security on 3 flaws in WebKit, the engine behind Safari. Apple said 2 of them meant maliciously crafted web content "may lead to arbitrary code execution," which means running an attacker's code. Code that breaks into the browser engine usually needs a second flaw to escape onto the rest of the device. One of those 2 fixes also credits a researcher with Anthropic, OpenAI's rival.
The fourth flaw sits in the kernel, the core of the operating system. Apple credits OpenAI alongside STAR Labs, a Singapore security firm, and said an app could use it to cause "unexpected system termination," a crash. Apple has not said any of the 4 were exploited in real attacks.
Researchers also used GPT-6 Cyber to write an exploit that strips protections from Lockdown Mode, the hardened setting Apple offers people at high risk of targeted attacks, Forbes reported. Calif chief executive Thai Duong still recommends it.
OpenAI launched GPT-6 Cyber on 5th November and rates it Critical for cybersecurity under its own Preparedness Framework, up from High for its previous cyber model. Only approved customers in its Daybreak Red program can use it. OpenAI confirmed to Forbes that its latest model found flaws Apple credited it for.
"I think things will become worse before they become better," Duong told Forbes.


















































