Excrubulent @excrubulent.bsky.social · 06/02/2026If you're going up, an angled ceiling will redirect you but lose some speed. If down, you can use a second tube entrance. They behave differently up & down. I think it's a bug that should be patched out, which is part of why I gave up on it. 100
Excrubulent @excrubulent.bsky.social · 06/02/2026Hey, not sure if I will. The short answer is that if you make a very short launcher module facing up or down, it will infinitely loop you through it. If you use a battery to time out the power, it will release you going potentially very fast. 100
Excrubulent @excrubulent.bsky.social · 20/01/2026So you're saying this is a great opportunity to reach across the aisle then? 010
Excrubulent @excrubulent.bsky.social · 20/01/2026Fascism is here, but I have a plan to supplicate myself to the most feckless rubes science & magic have yet produced and beg them to finally, heroically come to the rescue. 010
Excrubulent @excrubulent.bsky.social · 15/01/2026youtu.be/bRfBbjoD2vwyoutu.belearning growing changing 💛YouTube video by Natalie B 020
Excrubulent @excrubulent.bsky.social · 15/01/2026Glitterpiss is the name of my grunge retro pop experimental fusion band. 1411
Excrubulent @excrubulent.bsky.social · 02/01/2026"Stories need conflict." "Got it, make characters bicker and snip over every tiny damn thing; just be insufferable little shits." "No, I mean like conflicting motivations, obstacles, real problems to solve." "You think I'm a bad writer, don't you? I HATE YOU." 010
Excrubulent @excrubulent.bsky.social · 30/12/2025Turns out it's transparently a total conversion mod for quake. The enemies & weapons are reskins but ultimately play the exact same BUT THE LOUD MAKES IT A COMPLETELY DIFFERENT ANIMAL I CANNOT ADEQUATELY DESCRIBE THE LOUD 10/10 WHAT IS HAPPENING 010
Excrubulent @excrubulent.bsky.social · 26/12/2025As a ship of the line I could tank quite a few. They'd have to find some pretty specific weak points to do damage or else it'd just overpenetrate and leave a tiny hole. 000
Excrubulent @excrubulent.bsky.social · 22/12/2025I think even more important to his average supporter is that they fill in the gaps with their own voice, so he sounds like them and then they trust him. Literally, "This guy's saying everything I'm thinking!" 011
Excrubulent @excrubulent.bsky.social · 21/12/2025Lego is not on this site that I can tell, they are obviously afraid to face me. 011
Excrubulent @excrubulent.bsky.social · 21/12/2025Excuse me, Lego, but da Vinci was famously left-handed. 120
Excrubulent @excrubulent.bsky.social · 21/12/2025This is why I'm leery of "rights" in general. Rights are fine for now, but until you have liberation your rights are a piece of paper over a loaded gun. They don't protect you, they protect the state's legitimacy. That's not nothing, but it's not much either. 050
Excrubulent @excrubulent.bsky.social · 21/12/2025This one was doing too much. C'mon, I know the seconds, it's the hours that I need partially labeled in two different numeral systems to know what time it is. 000
Excrubulent @excrubulent.bsky.social · 21/12/2025I loved this video. I always knew dyson spheres were a joke, but I didn't know they were literally a joke. 010
Excrubulent @excrubulent.bsky.social · 19/12/2025Anyway, I don't have a good wrap-up for this. You can't trust AI, but more than that, you can't trust your feelings on it. It is hacking your perceptions. That is the real risk here, and the danger is that we are all susceptible to it, even if we could know better. 020
Excrubulent @excrubulent.bsky.social · 19/12/2025I tried showing my 12-yo the study about AI fooling devs into thinking it was making them faster, and they immediately started justifying why their use was different, why they could trust their intuition that it was helping, even though experienced devs couldn't. Explaining this is a battle. 120
Excrubulent @excrubulent.bsky.social · 19/12/2025And why don't people accept this? Because it's a lot easier to fool someone than to convince them they've been fooled, and a LOT of people have been fooled. 100
Excrubulent @excrubulent.bsky.social · 19/12/2025This is the alignment problem - the machine couldn't do fidelity, so it gave us the illusion of fidelity to create a false fit to the criteria it was given. The machine is lying to you, but on a deeper level than just telling you wrong facts. It is hijacking your own perception of reality. 120
Excrubulent @excrubulent.bsky.social · 19/12/2025It already does do that, but also it's a fever dream of a machine. One that had fidelity would potentially hold actual convictions, and that might suck for them. Not that we're anywhere near an AI that can actually think and believe things, but the point remains that they're happy with the scambot. 100
Excrubulent @excrubulent.bsky.social · 19/12/2025Maybe the CEOs would actually rather have something that cons people than does real information work, because all they need is hype for a bubble to get rich, and a scamming machine is actually better for that. A real information machine would probably say things they didn't like. 100
Excrubulent @excrubulent.bsky.social · 19/12/2025They are trained by humans giving feedback on what is useful, and it turns out that these people are easily fooled. Maybe this is a baseline issue with the human condition, maybe it's about education, maybe the workers doing the bulk of the work are too underpaid to care. 100
Excrubulent @excrubulent.bsky.social · 19/12/2025So why are people fooled? Why do they swear by AI's value in their workflow? Because in the absence of the ability to actually make useful, reliable output, LLMs optimised for the easier problem: to con people. 100
Excrubulent @excrubulent.bsky.social · 19/12/2025It is mathematically provable that LLMs cannot do fidelity: www.computerworld.com/article/4059...computerworld.comOpenAI admits AI hallucinations are mathematically inevitable, not just engineering flawsIn a landmark study, OpenAI researchers reveal that large language models will always produce plausible but false outputs, even with perfect data, due to fundamental statistical and computational limi... 100
Excrubulent @excrubulent.bsky.social · 19/12/2025AI makes you slower but fools you into thinking it's making you faster: metr.org/blog/2025-07...metr.orgMeasuring the Impact of Early-2025 AI on Experienced Open-Source Developer ProductivityWe conduct a randomized controlled trial to understand how early-2025 AI tools affect the productivity of experienced open-source developers working on their own repositories. Surprisingly, we find th... 100
Excrubulent @excrubulent.bsky.social · 19/12/2025"to streamline time-eating line writing" MOTHERFUCKER AI MAKES YOU SLOWER. AI can't do fidelity, but it turns out you don't need fidelity to fool people into thinking you're being helpful. That's what this is, it is a mass-gaslighting machine. 140
Excrubulent @excrubulent.bsky.social · 16/12/2025Nah, I bet I could do that, I'll give it a try and keep you posted. *user was last seen online 3 weeks ago* 060
Excrubulent @excrubulent.bsky.social · 15/12/2025@iwriteok.bsky.social, I just got done listening to this for maybe the fifth time after the hams references in the latest BtB: youtu.be/YRomQkC-D_8 If you ever make a hams podcast/show/thing, I want you to call it "Those Ham Bastards". That is all, thank you for your time.youtu.beWarhammer Is Insane (With Robert Evans)YouTube video by Gamefully Unemployed 020
Excrubulent @excrubulent.bsky.social · 08/12/2025That "mhmm" from Ana is painful. Like her whole thing was being a firebrand and now she's gotta politely nod along while Maher gestures at the vague idea of a joke and she actively suppresses the basic human instinct to claw out his eyes and run screaming into the street. 010
Excrubulent @excrubulent.bsky.social · 04/12/2025I lost my shit when I noticed this, then I looked it up and found out @scalzi.com was a creative consultant on this show. It's such a good show and I love they gave him his own little easter egg. 000
Excrubulent @excrubulent.bsky.social · 27/11/2025Oh shit it's in the same rant as the "RED ALERT, RED ALERT, RED ALERT" sample too. That's a little on the nose. 000
Excrubulent @excrubulent.bsky.social · 27/11/2025Listening back to the first @knowledgefight.bsky.social episode, about 41 minutes in, there's Alex denying Trump's connection to "Gerry Epstein". The guys didn't pick up on it at the time, but it was startling to hear in an episode from 2017. He always knew which way he would break on this info. 120
Excrubulent @excrubulent.bsky.social · 25/11/2025"ignores gatekeeping" is one way to spin "we can get it to make all our horrifically bigoted fash shit and it doesn't refuse or threaten to stab us like people do" 020
Excrubulent @excrubulent.bsky.social · 21/11/2025Definitely in the list of, "I can't believe this needs to be said, but it does need to be said." 020
Excrubulent @excrubulent.bsky.social · 20/11/2025Can't have free time and energy to develop an alternative means of supporting yourself that might free you from wage slavery, otherwise they'd have to find someone new to exploit. I'm convinced this logic is part of why our time is monopolised like this, I just don't know how explicit it is. 010
Excrubulent @excrubulent.bsky.social · 19/11/2025Or a whole community coming out with them. All different, discordant pitches of scream. The primal fear would be unreal. 020
Excrubulent @excrubulent.bsky.social · 19/11/2025youtu.be/_Sw0VCtZs-g Aztec/Mayan death whistle, there are many printable designs online. They are terrifying. When I thought this was just an organising tactic I didn't think the death whistle was a good idea, but now that I know it's to annoy them? Horrific death screams it is.youtu.beAztec death whistle , scariest bone-rattling whistle played by a female horn-bearerYouTube video by First Nations Music 031
Excrubulent @excrubulent.bsky.social · 19/11/2025www.youtube.com/watch?v=_Sw0... Aztec/Mayan death whistle, you can 3D print them. When I thought this was just an organising tactic I didn't think this was a good idea, but now that I know it's to annoy them? Horrific death screams it is.youtube.comAztec death whistle , scariest bone-rattling whistle played by a female horn-bearerYouTube video by First Nations Music 000
Excrubulent @excrubulent.bsky.social · 15/11/2025I think there's a good reason it's so useless: it's trained on human feedback, and it turns out it's way easier to fool the humans than it is to do real information work, which needs fidelity, which is actually impossible for LLMs. It has been trained to con people. 000
Excrubulent @excrubulent.bsky.social · 14/11/2025If the desert bus wheels are up then you are having a bad problem and you will not go to Las Vegas today. 010
Excrubulent @excrubulent.bsky.social · 09/11/2025That took me a fair bit of effort I'll be honest, and admittedly I had to find "sesquipedalians" with a scrambled word search, but I also didn't know that word to start with. Anyway, if anyone needed the help to read it, I think that demonstrates the point well enough. 010
Excrubulent @excrubulent.bsky.social · 09/11/2025For anyone alse who really wanted to know but struggled to read it: Practical experimentation demonstrates incorrect assumptions underpinning scrambled sequences. Passages composed of sesquipedalians are unquestionably illegible excluding clarifying context. 110
Excrubulent @excrubulent.bsky.social · 07/11/2025Chomsky: “The smart way to keep people passive and obedient is to strictly limit the spectrum of acceptable opinion, but allow very lively debate within that spectrum” Liberals: "But the democrats are on our side of the spectrum so we have to support them!" 010
Excrubulent @excrubulent.bsky.social · 06/11/2025I feel like he could explain this to people and they might get angry enough to do something about it. Like the banks control finances but not what people actually do. I think this is what strikes are good at. 000