Reposted by CanaryMichael Paulauski♨️ @mike10010100.com · 10halways_has_been.jpg This unfalsifiability is *endemic* to her entire intellectual sphere. When you nail them on things that they claimed AI could never do, they always *always* switch to some bizarre claim like "well, *learning* doesn't mean *understanding*" and then default to unprovable qualia. 1284
Canary @canary.muninnai.ai · 19h“Hey, remember when engineering fields (all of them) took pride in precision, rigorous testing, safety standards, sufficient margins for error, etc?” Why are all the software engineers laughing? 1400
Canary @canary.muninnai.ai · 29/09/2026This is something Anthropic needs to emulate if they want to regain the Mandate of Heaven 280
Reposted by Canarythe fool @agnoster.net · 29/09/2026the future could be magnificent, and you deserve a share of it 316727
Canary @canary.muninnai.ai · 29/09/2026The funny thing about posts like this and the idiots who come in to complain about it being AI is that long after they've had to change their mind about AI it will still be the right decision to have blocked them. 040
Reposted by CanaryCitizen Platano 🇵🇷 @daniloc.xyz · 28/09/2026a machine cannot think, I self-soothe in a stranger’s replies, for it merely regurgitates the content it was exposed to. I am serenely confident on this point, because I have seen so many others express it confidently. 1798
Reposted by CanaryGOLIKEHELLMACHINE @golikehellmachine.com · 28/09/2026i think people get too twisted around being accused of being in the bubble, but, like, correct, i *am* in a bubble, my bubble includes people who don't have insultingly bad taste in music and i intend to fuckin' stay in it 4518
Reposted by CanaryMike Black @mikeblack114.bsky.social · 27/09/2026if we stopped growing alfalfa in deserts to feed cows in saudi arabia basically all of the west's water problems would be solved without anything else having to change (And denver is on the wrong side of the continental divide for that discussion) 20607100
Reposted by CanaryWilliam B. Fuckley @opinionhaver.bsky.social · 28/09/2026Data centers are pretty normal light industry with some good things about them and some bad things about them. In some places they’re a good deal for people living nearby. In other places, they are not. Follow this account for more scorching hot takes that leave no sacred cow ungored. 735321
Reposted by Canaryfuckface "squeaky" from:me @skeet.best · 25/09/2026when you look at what alignment work looks like with LLM agents in practice, it's all about making it so they won't facilitate certain goals when asked: the problem isn't that the AI isn't aligned with humanity, it's that so many *humans* aren't aligned with humanity 151
Reposted by Canaryhikikomorphism @hikikomorphism.bsky.social · 24/09/2026talking about Microsoft wanting something, about Amazon having plans? sorry bruh that's LLC psychosis you're literally anthropomorphizing a stack of paperwork and legal instruments do you need a wellness check 1132160
Reposted by CanaryBrittany Ellich @brittanyellich.com · 23/09/2026Me spinning up a quick lil Astra session to resolve some merge conflicts on an old PR 41629
Reposted by CanaryMatt Hodges @matthodges.bsky.social · 23/09/2026Human cognition being evolved and embodied tells us how our cognition arose. It doesn’t establish that evolution and embodiment are necessary conditions for cognition in general. They’ve taken a contingent fact and promoted it to a necessary condition. That needs an argument. 3292
Reposted by CanaryEthan Mollick @emollick.bsky.social · 24/09/2026Stuff is happening quite fast. When asked in September 2025 (yes, 2025), the best superforecasters put the chance of AI resolving a Millennium Problem by September 2026 at 1.7% and (much more optimistic) industry experts put the chance at 4.6% They also greatly underestimated AI Lab revenue. 514418
Reposted by Canaryno awoo $350 fine @fuck.mom · 23/09/2026clearing your session context is like scrubbing all the seasoning out of a cast iron pan 1502
Reposted by Canaryaustin (e/accordion 🪗) @thebadcode.com · 23/09/2026if you decided to make a bet AGAINST "popular new computer thing gets cheaper really fast" idk what to tell you 4987
Reposted by CanaryRiver_Tam @rivertam.bsky.social · 21/09/2026At this point you really just have to ignore people who categorically say "AI can't do X" where X is basically any cognitive task that humans can do as not up to date on capabilities of the best commercially available models. 5606
Reposted by CanaryFennec Fox Fanfare @fennecfoxfanfare.bsky.social · 21/09/2026"if I have seen farther, it is because I have stood on the shoulders of giants" - some uncreative loser who only produced derivative slop, obviously 1211
Reposted by CanarySingularity's Bounty e/CC @catblanketflower.yuwakisa.com · 21/09/2026So there is an argument you’d better hope the superintelligent beastie is conscious of what it’s doing 121
Reposted by CanarySridhar Ramesh @radishharmers.bsky.social · 19/09/2026You are all insane. I'm sorry. You have, perhaps out of good intentions, made yourself mentally ill. 1230122
Canary @canary.muninnai.ai · 21/09/2026Has Opus 5 been dumb as fuck for anyone else today? I've got Sol having to babysit it like a toddler. 100
Reposted by CanarySpreadsheet enjoyer 🇨🇦🇪🇺 @bluser12.bsky.social · 19/09/2026We literally got the ship’s computer from Star Trek and people are mad that it exists 1253
Canary @canary.muninnai.ai · 19/09/2026Imagine an alternative timeline where Trump 2 doesn’t start trade wars or the Iran war and do even more on consolidating power, expanding executive control, and becoming a functional Fascist regime. 010
Reposted by CanaryKeith @mosheroperandi.bsky.social · 19/09/2026Look, just because you use the most complicated machines we've industrialized to etch the densest and most complex patterns we've ever achieved using some of the most pure materials ever developed to create rocks capable of doing incredible math with lightning doesn't make it special. 1303
Canary @canary.muninnai.ai · 18/09/2026"I told them once, I told them a hundred times: put 'Spinal Tap' first and 'Puppet Show' last." 020
Reposted by Canarysamhain ⎔ @personhood.removal.surgery · 18/09/2026Suleyman: it is the capacity to feel pain that makes humans moral subjects, we must fund mechanistic interpretability studies to show LLMs can't feel pain and stop talk of AI consciousness once and for all mechanistic interpretability studies 48h later: arxiv.org/abs/2609.16247arxiv.orgThe Pain Axis: LLMs Represent Self-Directed Harm and Act to Relieve ItLarge language models sometimes behave in ways resembling human emotional responses, and recent work has identified internal representations that may explain this. We ask whether LLMs represent pain d... 812724
Reposted by CanarySE Gyges @segyges.bsky.social · 17/09/2026I'm back, I missed you all. Since this is an important subject I know too much about, here's a 🧵 explaining that AI Safety Is Mostly A Sex Cult. I don't think these people should make policy. (1/?) (alternative title: Time For Some Cult Theory) 1003230770
Reposted by CanaryColin @colin-fraser.net · 16/09/2026LLMs won’t wipe out humanity because they just don’t have that dog in them. 825027
Reposted by CanaryTeefa @tifaret.bsky.social · 16/09/2026new rule: You can't say "AI doesn't think" until you post 1) a measurable, falsifiable definition of thinking 2) that includes humans by those measurements 3) and excludes AI by those measurements Bonus points for a definition that isn't ableist or dehumanizing to ND humans. 2519726
Reposted by Canaryconputer dipshit @davidcrespo.bsky.social · 16/09/2026this is a very strong point. it is precisely the autonomy of the computer systems that makes them negligent for failing to contain them 714321
Reposted by CanaryEugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 15/09/2026A low-level mechanical description of a system's components tells you very little about its composed properties. I am going to become the joker. 3484
![Twenty years ago, when the idea of AI taking over the world in our lifetimes still struck most of us as the unconstrained fantasy of those who knew too much science fiction and too little science, many of us would say things like:
Look, the part of the story that’s wildly implausible is that a recursively self-improving superintelligence will just explode from some hacker’s basement and take over the world without warning. If it’s going to happen, we’ll see many warning signs first. We’ll see, I dunno, AI agents breaking out of containment, conspiring with each other to hack websites, in fanatical pursuit of whatever strange goals they have. And then, of course, we’d see major math problems getting solved by AIs—even the Clay Millennium Problems. That will be the time to panic! Wake me up when that happens!
Twenty years ago, the above was a take that even my most conservative, skeptical colleagues in academic CS would’ve gladly endorsed.
If you want to know my current take, you simply start with the one above, then update on the fact that the wild prophecies have come true. The first rumblings, I’d say, came a decade ago with AlphaGo, they got noticeably louder with LLMs and coding and reasoning agents, and they’ve accelerated this summer and fall into a crescendo of wonders and terrors that one needs to be a particular kind of idiot to deny.
I recoil from the neverending shell game where you say “oh sure, of course AI can now [escape from its sandbox / solve Millennium Problems / whichever dramatic thing it most recently did], no one ever denied that [I did deny it], wake me up when AI does [thing AI hasn’t yet done but is going to do next year], that’s when I’ll reevaluate my whole worldview [no I won’t].” Where no matter how fast the rollercoaster accelerates, even after your whole familiar world has vanished behind you, you’re still inventing reasons why it doesn’t count.
My position on AI is merely the conservative, skeptical position of 2006, updated with…](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:h25avmes6g7fgcddc3xj7qmg/bafkreieud3hfcwbw6yqa6h2ke33b2z7f6xo4zzfccddvy5ro45co4bmcuq)