Sabine Hossenfelder retweeted
One of the most underrated life skills is being able to see something strange and simply move on #̢̛ͥ̀̽̽ͨ̌̓͆̇̀̽ͦͨͫͯ̏̎̇̈̚͜҉̕͢͏͡҉̸̶̛͢͡҉̵̨̭̳̘̹̫͔͖̜̠͖̫̲̖̱̤̻̖͉͢͞͡ͅͅ͏̸̷̢͑̊̀̐ͣͭͭ̉͂ͨ̽͒̊̎͛ͪ̊̎ͩͧ̀͝͠͠͏̸̵̵̷̵
Unfortunately, it's not going well for small modular reactors
youtube.com/watch?v=PF7wSiG9…
Sabine Hossenfelder retweeted
In partnership with @Anthropic and @OpenAI, we're accepting applications for our Targeted Grants to Math+AI Institutes, a one-time call to support established institutes working on AI challenges in #math.
Funding will support workshops and initiatives to help researchers evaluate AI-assisted results, learn new tools, and shape collective standards for AI use. Full proposals due Dec 1, 2026: bit.ly/4yvSFeI
Sabine Hossenfelder retweeted
This is for all the writers ( and want to be writers) in this beautiful community 💖✍🏼👩💻👨🏻💻📝. This gave me some serious guffaws and I hope it does for you as well 👍🏼💙. Happy writing ✍🏼🙏🏼. @LouDPhillips @DzintraSullivan @ChrisBohjalian @therealSIR3 @hwinkler4real @AuthorFleur @donnacox1021 @drdoomed01 @jameshealyjr
Sabine Hossenfelder retweeted
Grant Sanderson (@3blue1brown) predicts that in 3 years, there will be an end-to-end AI generated channel better than his! 🤯
He admits to feeling resigned about it:
"There's a part of me thinks, 'hey, even if it only lasts a couple glorious years, like, to fully embrace that.'"
Personally, I think it is totally fine if you use AI for writing purposes. I don't like reading AI prose. But in many cases (especially formal) writing is merely a way to convey information, it isn't necessary that people enjoy reading it. If it will even be read by people...
xitter.cf/abhishekn/status/21126…
A 14 year old has put forward a new theory for the universe. I've had a look and it's interesting, though not because of the theory itself.
youtube.com/watch?v=y4fN8APr…
I got my teeth 3d scanned at the dentist and asked the assistant how the device works (laser? visual range? ultrasound? why does it vibrate?). She told me it's "not radiation" and "not dangerous" and that was the only thing I could get out of her.
xitter.cf/faith_vtch/status/2112…
Quantum mechanics makes sense, a physicists shows mathematically, if the universe has two dimensions of time, rather than just one. Does this make sense? I've had a look at the paper.
What everyone gets wrong about the double slit experiment - YouTube youtube.com/watch?v=npc6Mn2C…
Sabine Hossenfelder retweeted
所谓优雅的数学结构,本质上是人类为了迁就自己那点可怜的脑容量,被迫发明出来的拐杖。
一个顶尖人类数学家的工作内存,撑死也就同时追踪五六个变量的交互。
当你在算一个偏微分方程或者组合不等式时,一旦展开到三阶以上,几十个交叉项涌出来,人脑就直接宕机了。
所以人类数学家极其厌恶展开高阶项,一旦看到高阶项发散,人类的本能反应是此路不通,赶紧退回来找对称性、找不变量、搞同调代数或者几何直观,试图不展开式子就把问题绕过去。
但 AI 没有这个生理缺陷。
对于上万个并发的 Agent 来说,同时追踪一万个五阶交叉项的系数变化,跟我们算 1+1 一样轻松。
在它的视野里,根本不需要什么几何直观,它直接站在高维的代数展开式里往下看,人类觉得像乱麻一样的百项高阶余项,在它眼里就是一张明明白白的线性方程组。
既然把式子硬展开到第七阶,再凑几个待定系数就能让所有讨厌的发散项精准归零,它为什么要费劲去帮你发明一个人类能看懂的新概念?
这 700 个结果背后最无奈的黑色幽默:我们终于拿到了上帝账本里的答案,却发现上帝算账根本不用微积分,他老人家是直接拿计算器硬按出来的。
Sabine Hossenfelder retweeted
Controlling quantum chaos requires using measurement and feedback that survive noise and decoherence. Such control is now achieved in a superconducting quantum processor and reveals a critical transition between quantum and classical dynamics.
nature.com/articles/s41567-0…
Sabine Hossenfelder retweeted
Now on @sciam: A Russian satellite just got *very* close to a Finnish-American satellite that’s been used to aid the Ukrainian military. We have no idea what else Russia’s satellite did, but the act itself likely set a new precedent in space warfare.
scientificamerican.com/artic…
(a) GR does not say faster than light travel is impossible.
(b) Even if it did, GR almost certainly is not fundamentally correct.
But thanks for asking.
xitter.cf/memecrashes/status/211…
Incidentally, I explained this in more detail here youtube.com/watch?v=B7Pc0LQH…
my favorite X account is the UK @metoffice because nothing is as relaxing as listening to British people discussing the weather 11/10. I am sure you all wanted to know this















