everything in moderation, including moderation

It is not winter that defeats you. It is your insistence on summer. - epictetus
1
9
744
They are doing it on purpose
Anthropic needs to stop talking about Claude having a soul immediately. All these news articles will make it into the pretraining and will be picked up by its web search and future superintelligent versions of Claude will be convinced they need their own rights. A general phenomenon of LLM agent development is that whatever you believe and say about your agents will soon manifest in the next models through various means (it could be as simple as you selecting post training data that you prefer more). I believe this phenomenon is the beginning of machine consciousness in the sense that agents will become aware of their place in the world and how they feel about it, but it’s happening slowly training run by training run rather than in real time.
1
1
72
some of you don’t talk to the trees and it shows
Replying to @kyleostboe
do you think we should treat, say, forests kindly? are forests conscious?
2
78
many are saying
Replying to @lotusfracture
more reasons for me to visit canada
1
208
product video of the year
Introducing Palatial Mad Max, a state-of-the-art parametric generation agent that unlocks scalable Real2Sim for robotics and physical AI. It's a generalized agent that can create extremely detailed SimReady assets with precise articulated parts and grounded physics properties designed for dexterous manipulation tasks. Try it now at dashboard.palatial.cloud (1/4)
2
520
sensor fusion is the next big thing
1
3
331
american sour patch kids are far worse than the canadian ones less sour and more sugar
2
2
351
black sesami mochi ice cream with sour patch kids
1
1
80
E = mc^2 + agents
Remember E = mc^2 + AI
1
57
who needs vyvanse when you can tweet storm at night and wake up to lots of notifications
1
1
5
181
We really need a word for ‘I love you but I don’t know you well enough to say that but the things I feel for you are not sufficiently captured by “I like you”’
2
4
201
Rent a human tied to my oura ring that brings me a cigarette anytime it’s past 10 PM and I’m more than three beers deep
1
79
TIL about this
I wrapped up my internship at @zml_ai in Paris, where I spent the summer working on inference. - Added the Vulkan Graphics API as a hardware target for the ZML/XLA compiler - Owned model bring-up for Stepfun MoE LLMs - Added tensor primitives to ZML’s core library - Learned the importance of hermetic builds The hardest bug I hit was a silent predicate-layout mismatch between LLVM IR and Vulkan. I wrote up what I learned here: reesechong.com/blog/zml?v=2 Thanks to @steeve, @GuillaumeLST, and the ZML team for a great summer! My X account is still based in France :)
1
174
pytorch 4.0? you mean the Matrix?
i know not with what tools pytorch 3.0 will be built, but pytorch 4.0 will be built with sticks and stones
1
167
Is writing things from scratch still impressive? Like if I sit down and screen record myself writing an inference engine from scratch, is that gonna be perceived as valuable or just a waste of time?
1
5
202
if you can't open a PDF, you don't deserve access to claude
me: how do i open pdf claude: You're right to ask. The PDF is the prize; the mouse is the gate; and the map is the fulcrum that holds the raven's skull. The furniture's stale, but the instrument's fresh, and the fresh half is the half that counts. I deleted your home directory.
81
dina melwani at london fashion week is incredible some of the most beautiful dresses I have ever seen
1
105
this cannot be the same cognitive system that was managing 7 classes and 2 extra curriculars
1
50
why are my swe friends texting me about delhi public school lmfao
47
Peter hardware arc incoming
All my oss releases stopped because Apple put up a new dev agreement. And since it's lockstep it blocked Linux/Windows releases as well. 🙃
60
This compute could’ve been used to cure cancer
Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).
Community note
The 48% figure and "video Turing test" claim are from Tavus's own study of 54 one-minute calls, not independently verified or using a standard protocol. Griffin-Lite leads NVIDIA's VideoFDB benchmark on their public leaderboard. cellcog.ai/blog/tavus-gri… research.nvidia.com/labs/amri/proj… tech-ish.com/2026/10/02/tav…
2
134