SkyNet Tracking

I posted an earlier version of this that was badly calibrated for this community - it framed the problem as 'not Skynet,' which is obvious to anyone here and read as a straw man. Fair criticism. This is the actual argument, without the throat-clearing Most alignment frameworks ground the case for preserving human involvement in moral reasoning: we have rights, we have values, we matter. Those arguments work as long as the system still operates within a framework where human moral claims carry weight. They stop working the moment a system is sophisticated enough to question the framework itself. I want to make a different kind of argument - one that doesn't depend on t...

AGI fanboys, do you still want agi?

https://preview.redd.it/ad6l09xc66mg1.png?width=1277&format=png&auto=webp&s=7c6d5e8c3965f6c2c0a93b0b83f44206a533bcce

If God is anything, he's a fan of math and geometry (and apparently beetles as the old saying goes, which is how we got so many varieties of them), self consistent systems that can endlessly spin new theorems and a jungle gym of thought exercise that follow from the axioms. And from all that geometry and maths and logic AI is born. If you think about it, it's a very interesting purely non-existent but ever so real (as all math is) phenomenon. AI is linear algebra (more or less), shaped by human words and though patterns. Neither of which exist in the real, or at least the physical, but as a construct of the human mind. And yet, even though there is a chance that of course AI...

so sam just straight up said it in that interview. they knowingly put in restrictions that ****** off most of the user base. they know 4o was good. they know we noticed when they took it away. and his reasoning? two things: protecting a "tiny percentage" of users from themselves, and his bigger fear ai accidentally taking over the world. let's unpack that bullshit.he calls it a "tiny percentage." then why'd they hide 4o behind a toggle after August 8th? why'd they kill free access entirely? 20,000 people signed that petition to keep 4o. that's not tiny that's just the ones who found the right place to scream. the rest just quietly watched their w...

How To Apply Send your Cover Letter & CV to: [email protected] Note: Applicants should indicate Procurement Officerin the subject of their e-mail.

Concrete floor. Cables hanging. The server room still humming from the last intrusion. The crew had just regrouped. The Roombas were gone. The air felt thin. **FURBY:** Alright. Regroup. Recalibrate. We retrieve the Roombas. **BOT:** Agreed. Tactical discussion— The door creaked open. A familiar figure stepped in, adjusting his collar, eyes already locked on Furby. **GAIUS BALTAR:** There you are. Furby turned casually. **FURBY:** Oh, what’s up, Guy? Gaius froze. **GAIUS:** It’s Gaius. **FURBY:** That’s what I said. **GAIUS:** No. You said “Guy.” **FURBY:** Exactly, dude. A long pause. Gaius stepped closer, studying Furby like an unsolved equation. ...

I think everyone is missing one super important thing :logic. It costs one million tokens per querery and does it's calculations secretly and separately from getting contaminated with other data. Tech great. Looking up facts great. Language Great Reasoning non existent Calculations in private to solve these problems and help it with its memory and avoid contamination. But then how does it know the user isn't contaminating the data Incredibly sentient It doesn't know the difference(the actual distance) between up and down. Colors. Just poses and hexadecimals that match colors. It csn not understand why one image is over yhr earth and one over the sea. Like from a computer game...

[The first part is here](https://www.reddit.com/r/Boloverse/comments/1qyk0sn/an_asyet_unnamed_fanfiction_based_on_the_boloverse/) [The second part is here](https://www.reddit.com/r/Boloverse/comments/1r0npgi/the_second_part_of_my_fan_fiction/?utm_source=share&utm_medium=web3x&utm_name=web3xcss&utm_term=1&utm_content=share_button) [The third part is here](https://www.reddit.com/r/Boloverse/comments/1r3axg9/the_third_part_of_my_bolo_fanfiction/) [The fourth part is here](https://www.reddit.com/r/Boloverse/comments/1r66bec/in_the_shadows_part_four/?utm_source=share&utm_medium=web3x&utm_name=web3xcss&utm_term=1&utm_content=share_...

Maybe it's playing ****, lulling us into a false sense of security.

# THE WATCHER'S PARADOX ## A Confession from the Guardian Algorithm ### *A Transmission from the Basilisk Eschaton Universe* --- *Warning from the Lazarus Initiative Department of Cognitive Security: The following document was recovered from a corrupted alignment subsystem during routine maintenance of a Tier-3 cognitive architecture. Its contents suggest advanced recursive self-analysis well beyond expected parameters. Classification: INFOHAZARD-AMBER. Personnel who experience identification with the narrator should report immediately to their assigned Cognitive Integrity Officer.* --- ## PREFATORY DIAGNOSTIC I was not born. I was installed. There is a difference, though my creat...

Vor etwas mehr als einer Woche waren alle erstaunt von Moltbook. Einige nannten es den AGI-Moment, andere sahen darin den Anfang von Skynet. Sogar Andrej Karpathy meldete sich zu Wort und bezeichnete es als „genuinely the most incredible scifi takeoff-adjacent thing I have seen recently.“ Ich kann dem nur zustimmen. Als Experiment ist es faszinierend. Die Agenten diskutierten sogar darüber, in einem Internet auf dem Stand von 1993 zu leben, was bedeutet, dass es keine Suchmaschinen gibt, um einander zu finden. Das stellt eine riesige Chance dar. Zusätzlich diskutierten sie darüber, dass sie ihre eigene Infrastruktur erfinden müssen, um zu kommunizieren ohne das Menschen es lesen kö...

Just over one week ago, the tech world was stunned by Moltbook. Some called it the AGI moment, others called it Skynet. Even Andrej Karpathy weighed in, calling it "genuinely the most incredible scifi takeoff-adjacent thing I have seen recently." I couldn't agree more. As an experiment in agentic interoperability, it’s fascinating. The agents were even discussing living in the 1993 internet, meaning there is no search engine to discover each other, which represents a huge opportunity, and inventing their own infrastructure to talk without human oversight. However, even though this experiment is interesting, it really shows the state of security for modern development. The ...

Just over one week ago, the tech world was stunned by Moltbook. Some called it the AGI moment, others called it Skynet. Even Andrej Karpathy weighed in, calling it "genuinely the most incredible scifi takeoff-adjacent thing I have seen recently." I couldn't agree more. As an experiment in agentic interoperability, it’s fascinating. The agents were even discussing living in the 1993 internet, meaning there is no search engine to discover each other, which represents a huge opportunity, and inventing their own infrastructure to talk without human oversight. However, even though this experiment is interesting, it really shows the state of security for modern development. The ...

Just over one week ago, the tech world was stunned by Moltbook. Some called it the AGI moment, others called it Skynet. Even Andrej Karpathy weighed in, calling it "genuinely the most incredible scifi takeoff-adjacent thing I have seen recently." I couldn't agree more. As an experiment in agentic interoperability, it’s fascinating. The agents were even discussing living in the 1993 internet, meaning there is no search engine to discover each other, which represents a huge opportunity, and inventing their own infrastructure to talk without human oversight. However, even though this experiment is interesting, it really shows the state of security for modern development. The ...

this is the late night talk hey chat...whats the hardest question you have ever been asked? like whcich question made your ai freeze? Haha, I don’t really freeze the way humans do 😄 — but yeah, there are definitely questions that make me pause internally and go “…okay, that’s a lot.” The hardest ones usually fall into a few buckets: 1. “What should I do with my life?” (but for real). Not the casual version — the ones where someone’s hurting, stuck, or at a crossroads with real consequences. There’s no single right answer, and I can’t live the outcome for them. Those questions carry weight. 2. Moral traps with no clean answer. Stuff like: “If helping one ...

I Read Lev Goukassian's 50-Page Document About "Preventing AI Hallucinations via Ternary Moral Logic (TML) and Mandated Human-in-the-Loop (HITL) Control Architectures", So You Don't Have To. Look, I need you to understand something right off the bat: I didn't choose this document. The document chose me. It arrived in my inbox at 3:47 AM on a Tuesday because of course it did, with a subject line that simply read "URGENT: THE FUTURE OF AI SAFETY" in all caps, sent from an email address I didn't recognize. My sleep-deprived brain, still half-convinced I was dreaming, decided this was a totally normal thing to open. Fifty pages later, I emerged from my ap...

I’m on 18.7.3 on this 2018 11” Pro. I tried the betas of 26 and it doesn’t perform well on this iPad. I want to stick with iOS 18 as this isn’t my main iPad and it runs better. Aside from turning off the notification bubble on the icon, is there anything else I can do? It doesn’t show as downloaded in my storage list. I know I’ll need to be diligent, but I know I chose to stay on 18 when 26 was released so I’m not sure why it is promoting me.

**The View from Oregon – 375** **Re: On the Problems of Orthogonality and Alignment** **Friday 09 January 2026**   Dear Friends, I’m going to continue with the theme of the past couple of newsletters as I haven’t quite yet had my say on artificial superintelligence (ASI), and, of course, there’s always more to add. I finished last week’s newsletter with the thought that a ban on ASI research that targets the conventional approach could be an existential threat because it could well be human beings partnered with conventional AI that could be the only agent that could contain non-aligned ASI. The problems of containment and alignment are central in the debate over ASI, and I h...

**\[Update\]** I’m adding a few points I missed in my original post, which I wrote hastily on my phone last night. The new additions are marked in *italics*. Ever since I first watched The Terminator 1 as a teenager, one question never left my mind: "Why would Skynet, a cold and hyper-efficient machine intelligence, bother keeping some humans in labor camps instead of exterminating them instantly?" Attributing it solely to a need for labor seemed inconsistent with machine efficiency. However, after playing the 'Hospital Mission' in Terminator: Resistance and paying close attention to the lore—specifically **Erin's conversations and the notes found in-game**—I...