Sabitlenmiş Tweet
Clockwork Whale
2.4K posts
@mmjukic please stop boosting fake third-worldist slop, this is the second time you've done it in a few days (last time was claiming crippling damage to a US aircraft carrier)
I'm not happy about this retarded war either, but that's why I need accounts like yours to keep their heads
English
Three weeks I've been telling copers that Iran will just bomb Saudi Arabia's East-West Pipeline and its infrastructure or ships on the Red Sea.
And here it is, as expected.
خبرگزاری فارس@FarsNews_Agency
🔴 اصابت موشک به پالایشگاه ینبع در قلب دریای سرخ منابع محلی دقایقی پیش از انفجار در پالایشگاه ینبع در قلب دریای سرخ خبر می دهند. سرمایهگذار اصلی این پالایشگاه، آمریکا است و این اولین بار است که دامنه آتش به دریای سرخ کشیده می شود.
English
sympathetic to Eliezer's position here because I think what we see with stuff like Breaking Bad or The Sopranos is that the median viewer is actually incapable of not identifying with an evil protaganist, simply because they're the protagonist, and then rationalising them as not-evil because of the identification
it's really visible in The Sopranos where you can see the writers becoming frustrated with the audience's refusal to see Tony for what he is and intentionally writing him to be worse in later seasons, and even breaking the fourth wall a bit with side characters explaining that he's bad. But it still didn't really work
English
@allTheYud @roryisconfused @robbensinger Not sure if you've watched Breaking Bad, but I definitely wouldn't characterize it as a morally deficient show in the way that its place in this conversation and the Dexter comparison imply.
English
I don’t know, it’s actually very depressing that a lot of people in the tech sphere genuinely have this view (which is somehow more reductive than the Hays Code) that fiction *must* be about a good person doing good things so as to set a clear example
Ambivalent Rory@roryisconfused
Silicon Valley’s most influential intellectual
English
@witte_sergei seems to be a really tiresome risk of being sympathetic towards Russia that you collect a bunch of dipshit third-worldist followers who expect you to agree with them on everything else too
English
Aaaaaaand people are calling me Jewish because I don’t think Israel’s blackmailing Trump with Epstein Island sex tapes.
English
I have some thoughts about the Iran War up over on the ‘stack. Link below.
English
@conan_esq realizing it was fundamentally a silly costume larp for nerds made me enjoy it more rather than less
English
When I was a hs freshman I remember my stupid normie chud gym teacher randomly walked by in a polo and khakis and saw me reading Dragonlance and listening to Hollenthon and had a whole conversation with me about Raistlin and With Vilest. That’s when I realize all this shit is gay
𝖕𝖆𝖙 🌙🗡️@dawnofpatrick
*taps the sign*
English
@teortaxesTex divine intervention from God to prevent the blasphemy of ASI
wonder how he'll get Sam
English
Uncanny that we are seeing multiple major AI labs in serious trouble. xAI basically imploded (Elon copes that it "was not built right the first time"), Meta continues to fail with its R&D, Anthropic got into trouble with the Government. None of that makes much sense.
English
@woke8yearold someone doesn't actually have to look bad for a malicious photojournalist to make them look retarded
there's all sorts of petty childish tricks they use on people they don't like, like publishing a photo that was taken mid-blink
English
"You really don't want to eat that apple."
Have to admit I'm the worst possible audience for this.
English
@teortaxesTex one thing is that both Hunter and Healer force user-message-last in the request and throw a 400 error if given a conversation with the assistant last. DS API has never done that, they always support assistant last with prefill
(I'm coping a bit because I don't like these models)
English
@clockworkwhale nah, they could, for a change
English
hunter does feel to have a similar vibe to what's on deepseek-Web, and sadly I can't say so far that it's stronger. Minor differences in reasoning style.
Healer seems to have worse general knowledge and not really frontier in multimodal understanding
AiBattle@AiBattle_
2 new Stealth models on OpenRouter Hunter Alpha: - "Hunter Alpha is a 1 Trillion parameter + 1M token context frontier intelligence model built for agentic use. It excels at long-horizon planning, complex reasoning, and sustained multi-step task execution, with the reliability and instruction-following precision that production agentic pipelines demand" Healer Alpha: - "Healer Alpha is a frontier omni-modal model with vision, hearing, reasoning, and action capabilities. It brings the full power of agentic intelligence into the real world: natively perceiving visual and audio inputs, reasoning across modalities, and executing complex multi-step tasks with precision and reliability"
English
@tszzl scully starting to seem like an absolute lunatic by still refusing to believe in ghosts in season 4
English
@LS_Keyes "you're a hypocrite if you think it's bad to give money to people inciting murderous racial hatred with fake stories"
bit silly man
English
Two years ago: Wow. Monetization is great idea, Elon.
Now: The Third World is better at gaming the system then us. Shut it down, Elon!
Lol. Lmao. You guys asked for this.
English
hmm maybe if it was fully upfront on turns like that
just thinking about how my blue collar family members talk about power bills, it's facially above board with kilowatt-hour pricing and they know "use more = more money", but somehow they're suspicious of it, see it as in some way less trustworthy than services that charge them a flat fee
English
@clockworkwhale I disagree. Hide the token stuff and just make it transparent the price of every turn. It makes sense to me
English
@Rudo1518568 @allTheYud The same way that anti-nuclear activists succeeded in killing civilian nuclear development but did nothing to stop nuclear weapons, you are going to be used to prevent the good uses of AI while not stopping any of the violent and oppressive uses
English
@allTheYud Just like left never countersignals those to their left, we should never countersignal those who are more ban/pause/block prone then we are.
English
It feels like there's a determined effort to end up on the worst timeline. Like how it was impossible to build nuclear power plants right up until they were going to be used for training AI that will wipe out humanity. You can only have the thing if it isn't nice.
More Perfect Union@MorePerfectUS
A New York bill would ban AI from answering questions related to several licensed professions like medicine, law, dentistry, nursing, psychology, social work, engineering, and more. The companies would be liable if the chatbots give “substantive responses” in these areas.
English
@benlandautaylor people's hearts weren't really into turning it in a new moral panic, because deep down they know social media psychosis is already a way bigger problem and the LLM variant is a mere sideshow
English
Whatever happened to "AI psychosis"? Did chatbot companies RLHF it away? Did it turn out to be a small handful of cases we can ignore? Did the public evolve antibodies? Did everyone just get tired of talking about it? It dropped out of the discourse and I have no idea why.
English
@illegal_lemur @eigenrobot yeah lol
but witnessing the degree to which they've been on their feelings the last week I am now 100% convinced they are not above it
timelines just got pushed back at least a couple years
English
@clockworkwhale @eigenrobot The idea of slow rolling the release of the cosmos-consuming machine god because of petty lib shit is pretty funny actually
English
happy singularity to those who celebrate
Arram in NYC@arram
Wishing you all a positive singularity.
English
@allTheYud @tszzl yeah, for the guy believing he's working to save humanity by posting infantile console wars content
English
Apparently Claude is currently struggling under a weight of new subscriptions. If you have an old ChatGPT sub you just cancelled that hasn't run out yet, consider putting all your weight on ChatGPT that it can handle, and sparing Claude when possible for a week or two.
English
@pmarca The question on everyone's minds is whether they'll continue feeling that way once the government is no longer chud-coded
Disguising outgroup-disgust as noble principle is the easiest thing in the world
English
Overheard in Silicon Valley: "Every single person who was in favor of government control of AI, is now opposed to government control of AI."
English