- Best of Twitter
- Posts
- Best of Twitter week of December 23-29, 2024
Best of Twitter week of December 23-29, 2024
Absolutely. In any hypothesis test between A and B about the working of a complex system, the right answer is invariably none of the above. Systems identification is a much better paradigm for neuroscience discovery; it allows us to efficiently explore huge hypothesis spaces.
— Surya Ganguli (@SuryaGanguli)
6:38 PM • Dec 20, 2024
As I’ve gotten older, it seems most wise to marry at 24-27 for women and 27-30 for men.
From my observations over the course of my life, people tend to wait a couple years before having children anyway so delaying marriage really delays child rearing until you are bumping… x.com/i/web/status/1…
— Sivori (@sivori)
3:33 PM • Dec 28, 2024
Similar story in THE NVIDIA WAY. Intel tried to recruit top engineers from 3DLabs in Huntsville, AL, but required relocation. Jensen swooped in, offered jobs at Nvidia, letting them stay in Huntsville with families. Don't let bureaucracy hinder talent acquisition.
— tae kim (@firstadopter)
4:44 PM • Dec 24, 2024
for the past month i’ve been desensitizing myself to notifications by having my phone send me random pushes
it hasn’t worked at all, but I do check my phone more and feel insane all the time
— nolen (@itseieio)
9:53 PM • Dec 24, 2024
An issue similar to the arbitrary p-value cutoff in some areas of science is the 5σ threshold for discoveries in some areas of physics. It's ultimately an arbitrary measure that became used, I think, just because they have so much data and it reduced the number of potential… x.com/i/web/status/1…
— Sabine Hossenfelder (@skdh)
7:15 AM • Dec 24, 2024
Why you shouldn't build your career around existential risk.
I feel weird writing this because the core of the argument is almost metaphysical for me. I believe that attention is the most powerful thing in the world and I have a very deep sense that whatever we pay attention to… x.com/i/web/status/1…
— Alexey Guzey (@alexeyguzey)
5:42 PM • Dec 23, 2024
People who are going to change the world: guzey.com/people/
— Alexey Guzey (@alexeyguzey)
8:42 PM • Dec 27, 2024
I made this graph in 2014, and a decade later I think I understand why open debate in economics declined: the peer review system took over, and the debate moved behind closed doors.
This trend was, moreover, not just confined to economics. Peer review killed open debate.
— joseph francis (@joefrancis505)
10:58 AM • Dec 23, 2024
AI can show emotion
— adi (@adonis_singh)
2:36 PM • Dec 23, 2024
^ thread
Unitree B2-W Talent Awakening! 🥳
One year after mass production kicked off, Unitree’s B2-W Industrial Wheel has been upgraded with more exciting capabilities.
Please always use robots safely and friendly.
#Unitree#Quadruped#Robotdog#Parkour#EmbodiedAI#IndustrialRobot… x.com/i/web/status/1…— Unitree (@UnitreeRobotics)
7:44 AM • Dec 23, 2024
^ i keep rewatching this video
it feels mildly fucked up to watch the medical advice given to mothers be completely different to almost opposite to that of the medical advice that medical practitioners actually follow when they themselves become mothers
like a clinician on our team straight up doesn't… x.com/i/web/status/1…
— zeta (@zeta_globin)
3:19 AM • Dec 23, 2024
Can interpretability help defend LLMs? We find we can reshape activations while preserving a model’s behavior. This lets us attack latent-space defenses, from SAEs and probes to Circuit Breakers. We can attack so precisely that we make a harmfulness probe output this QR code. 🧵
— Luke Bailey (@LukeBailey181)
5:54 PM • Dec 13, 2024
Reply