Congratulations to the @Google Networking team to yet another SIGCOMM Networking Systems Award. And congrats to Amin Vahdat, Steve Gribble, Urs Hoelzle and probably a few more that I am missing to their second award.
Google continues to have the best networking infra of any compute provider.
Love this interview of Amin Vahdat from Google. Getting deep into the engineering details. And the Acquired Podcast interviewers! 👍 https://t.co/PH3msbP5sT
Our internal data shows Claude is accelerating AI development—a possible path to recursive self-improvement, or AI autonomously building a more capable successor.
It’s happening faster than we thought, and the implications deserve greater attention. https://t.co/OVVPJO7VQx
Anthropic's Chief Economist:
"The gap between AI capability and what people actually understand about it is enormous."
In 30 minutes he explains how millions of workers are already pasting sensitive company data into AI tools without knowing where it goes.
Watch the interview, then save full privacy setup below 👇
Complaining about things is one, but fixing them myself is another
So I added a [ 🗽 Not woke ] filter to https://t.co/kSbsCmuvBO
"woke" is a bit of a cringe term now, but calling it "Not ESG" would be too difficult
I scraped ALL hotel chains and sites for stuff like "climate change", "sustainability", "ESG", "carbon neutral" and tagged all of them
So now you can AVOID hotels that:
- limit your AC to 25C/77F (common in Europe)
- skip cleaning you room altogether (last 2 hotels I stay in Europe now)
- have no amenities (again I just stayed in them), and other useless stuff that does nothing to save the world at all
- reduce shower pressure to save water so you can't clean yourself
- charge you extra for your CO2 emissions every day you stay (???)
In fact all of this does the direct opposite of make the world better: it makes everyone hate their stay, sweat and have bad sleep, and be grumpy and unproductive all day
Avoid them with my site Hotelist and make the world better! :D
What is mid-training?
The stage between pre-training and post-training
A base model is continued on a smaller, curated data mixture chosen to strengthen capabilities that the original pre-training run undercovered, such as multilinguality, domain knowledge, or long-context extension.
It usually keeps a pre-training-like objective, but uses higher-quality or more targeted data so later instruction tuning, preference tuning, or RL can shape behavior on top of stronger capabilities.
Learn more here: https://t.co/WhpYkyGlv8
Super excited to open source Flashdreams, acceleration library for world models! No restrictions on usage, Apache 2 license.
World models are the next frontier for video generation and paramount to Physical AI.
Flashdreams yields out of the box speed up across world models
Sergey Brin joined Rocky Yu for an unscripted fireside chat
On topics like superintelligence: or how specialized scientific models for tasks like fusion reactor management and protein folding are converging into general purpose models like Gemini.
In TripoSplat, we sort of figured out how to “model the distribution of 3D Gaussians” instead of “predict 3D Gaussians given fixed structures”.
The Gaussian centers are sampled from a distribution which can be optimized end-to-end using rendering supervision!
Learn more👇
Anthropic engineer:
"You can build 5 assistants in one afternoon. Each one handles a task you've been doing manually every single day"
in 45 minutes he shows exactly how to do it from scratch, step by step
most people are still doing this manually
watch the session, then save the guide below
Took one of our @personaaiinc Gen 1 humanoids for a stroll outside our offices in the Houston heat a couple weeks back. Testing out some new @UnderArmour treads. Also stay tuned for our upcoming Gen 2 launch!
Building apps has never been easier.
With Sites, Codex can turn your work, ideas, and plans into an interactive website or app your team can explore, use, and share with a URL.
Rolling out to Business and Enterprise plans, before expanding more broadly.