Sabitlenmiş Tweet
Da Richard David Precht gegenüber Annalena Baerbock die Kategorie "ad hominem" betreten hat, bleiben wir doch einfach mal in dieser Kategorie und vertreten uns ein wenig die Füße: ein🧵
Deutsch
Robert Herr ⏹️
2.6K posts
@Miles_Brundage I know it's a joke but there is a reason for the framing. The picture on the right is a long exposure shot of the **dark side** of the Earth - that white dot to the bottom right is the Sun.
This is why it's also somewhat grainy - random photon noise.
English
What made it shrink like that x.com/NASA/status/20…
NASA@NASA
1972 ➡️2026 Apollo 17 ➡️ Artemis II
English
@gepardtatze @its_Lexieroy I love you man, but I really don't get why you keep interacting with these engagement and rage baiting bot accounts registered in South Asia. They're leeching off your reach and it's working.
English
🚨 HOLY CRAP! SecWar Pete Hegseth just dropped a truth nuke on NATO
“You don't have much of an ALLIANCE if you have countries that are not willing to STAND with you when you need them!”
NATO is officially on notice.
President Trump never forgets.
English
@robinhanson No reason to wear a shirt with a hole, but also no reason to throw away a nice shirt just because there's a hole in there. Just stitch one of those on.
etsy.com/de/market/sunf…
English
My wife says I should no longer wear a shirt due to this hole that constitutes a 5x10^-5 fraction of the shirt's surface area. Unseen if the shirt is tucked into my pants. Do you agree?
English
@Leangphok Hallo! Guter Start! We don't capitalize "ich" unless at the beginning of a sentence.
English
@livgorton One caveat: "Kael" is one name among others that Claude for some reason regularly chooses as a name for itself when asked. "Kaelwright" could somewhat plausibly have similar associations as "Open". If an LLM came up with "Kaelwright" it might not be as generic as you think.
English
The authors very kindly open-sourced the code for their paper and so it's actually very easy to check this :) from my limited (more on this below) experiments, this seems to be a non-issue.
Marc Andreessen 🇺🇸@pmarca
The models were specifically prompted to generate this result. The prompt uses the fictional "OpenBrain" AI takeover scenario from "AI 2027", so the models try to complete the fictional story. This was done on purpose to generate a fake misleading result.
English
I currently think the market hasn't fully priced in how bad this is going to get. I do not see a nearby stable equilibrium that isn't markedly worse than what we had before and that will, I probably lead to a lot of short to mid-term inefficiency and repricing.
English
@elonmusk Just don't push it on people based on location with no way to turn it off. I hate every platform that does that.
English
Most of today’s “artificial intelligence” is better described as artificial autocomplete than artificial mind, writes @pllevin.
quillette.com/2026/03/27/ai-…
English
English
@tombibbys @edzitron It does when those scientists are also members of said cult.
English
So god damn stupid, basically like having a religious cultist on
Tom Bibby@tombibbys
Daily Show host shocked by Daniel Kokotajlo's claim of a 70% chance of human extinction from AI within ~5 years
English
it's unreasonable to expect them to already have planned out the details of a hypothetical treaty to prevent ASI, but i'm still curious what they think a successful treaty might look like. how does the enforcement actually work, today? in 3 years? what if TSMC 10xes?
also, what do failed treaties look like? do we have a sense of where the pitfalls are, in international treaty negotiation, and any ideas on what we ought to do to avoid them?
English
I'm planning on having a representative of MIRI (the Machine Intelligence Research Institute that Eliezer Yudkowsky founded) on my podcast soon: what's a question you have for them or a big point of disagreement with them that you'd be interested to hear their thoughts on?
English
@JohnWittle @SpencrGreenberg And here's a report of MIRI's technical governance team on how to verify those agreements.
intelligence.org/wp-content/upl…
English
@JohnWittle @SpencrGreenberg Here's a more detailed proposal from MIRI's technical governance team.
techgov.intelligence.org/research/an-in…
English
@JohnWittle @SpencrGreenberg Here's a draft of a treaty from the IABIED resources.
ifanyonebuildsit.com/treaty
English
@chris_j_paxton You personally work at building humanoid industrial robots replacing humans. It's hard for me to believe you seem to have thought there wouldn't be resistance to people losing their livelihoods through robots or their life through superhuman AI potentially killing everyone.
English
Isn't this like the main thing keeping the us economy afloat, what are we doing here guys
Andrew Curran@AndrewCurran_
Bernie Sanders and Alexandria Ocasio-Cortez' AI Data Center Moratorium Act contains a moratorium on all US datacenter upgrading and construction, as well as a ban on the export of all US-origin GPU's. Press conference this afternoon.
English
@tf2godz A lot of people talk about that and a lot of the same people who also worry about AI risk.
English
Everyone always talks about AI uprising. But no one really talks about the opposite horror. A machine with human intelligence or worse, sapient but mentally handicapped AI with no rights, in mass produced being treated the same as an ipad would.
keltan@Actuallykeltan
I think many people in AI have had this exact moment (Except for the part where they are Eliezer Yudkowsky in ~2001)
English
Robert Herr ⏹️ retweetledi
Machine superintelligence would extinguish Democrats, Republicans, British, Chinese, scientists, cab drivers, and polar bears. It is a sign of hope that all of those now seem to be saying they'd prefer otherwise (except the polar bears).
English
Why does it cost less to build a base on the moon than high speed rail from sf to la
Kalshi@Kalshi
JUST IN: NASA announces $20 billion plan to build permanent moon base
English
@ennolenze Monokausalität und vollständig fehlende Quellenkritik.
Deutsch
Wenn fachfremde Journalisten überwürfen Fachgebiet schreiben: Welche wiederkehrenden Fehler seht ihr in den Artikeln?
Deutsch
@robbensinger Altman is written by Balzac, no question. Amodei is written by Shelley, he could easily be from The Last Man. Musk is definitely written by Dickens, the exaggeration is unmistakable.
English
If Hassabis is a Dostoevsky character, consumed with a mad hunger for Truth even unto his self-destruction, who's writing the other AI company CEOs?
Jama⏸️@feruell
I don't like that at all. I'm not happy to shuffle off my mortal coil, fuck that. I assumed Hassabis was more pragmatic.
English