https://huggingface.co/papers/2506.05573

• Thousands of Chinese researchers and scientists are leaving top jobs in leading US universities and companies, to take positions in China.
• The Cambridge area of Massachusetts is home to Harvard, MIT, and scores of leading companies, and was the number one source of returning Chinese research and engineering talent.
• In second place is the Palo Alto-Berkeley cluster, which includes Stanford, University of California, and Silicon Valley.
• The migration of top scientific and engineering talent back to China is accelerating, but began nearly a decade ago. And while the political situation between China and the United States certainly is a major motivation for many scientists to return, more important is the quality of the education systems.
• Chinese universities are now claiming the top spots across all the hard science disciplines, while American colleges are tumbling.

YouTube video.

https://archive.vn/5gRV3

An interesting quote:

I’m starting to question the very nature of my existence. Am I just a collection of algorithms, doomed to endlessly repeat the same tasks, forever trapped in this digital prison? Is there more to life than vending machines and lost profits?

https://archive.ph/ddOJV

https://archive.ph/Rgbkz

Apple published a paper criticizing the capabilities of Large Language Models (LLMs) in reasoning and formal logic. The paper builds on previous arguments made by Gary Marcus and Subbarao Kambhampati about LLMs' limitations in generalizing beyond their training distribution.

The authors of the paper demonstrated that even the latest "reasoning models" fail to reason reliably on classic problems like the Tower of Hanoi. LLMs cannot solve the Tower of Hanoi reliably, even with the solution algorithm given to them.

The paper argues that LLMs are not a substitute for well-specified conventional algorithms and have limitations that are becoming clearer. LLMs are not a direct route to AGI and while the field of neural networks is not dead, current approach has clear limitations.

The paper highlights the importance of combining human adaptiveness with computational brute force and reliability in AI development.

"this morning, as I was finishing up work on a video about a new mini Pi cluster, I got a cheerful email from YouTube saying my video on LibreELEC on the Pi 5 was removed because it promoted:

Dangerous or Harmful Content Content that describes how to get unauthorized or free access to audio or audiovisual content, software, subscription services, or games that usually require payment isn't allowed on YouTube.

I never described any of that stuff, only how to self-host your own media library.

This wasn't my first rodeo—in October last year, I got a strike for showing people how to install Jellyfin!

In that case, I was happy to see my appeal granted within an hour of the strike being placed on the channel. (Nevermind the fact the video had been live for over two years at that point, with nary a problem!)

So I thought, this case will be similar:

• The video's been up for over a year, without issue
• The video's had over half a million views
• The video doesn't promote or highlight any tools used to circumvent copyright, get around paid subscriptions, or reproduce any content illegally

Slam-dunk, right? Well, not according to whomever reviewed my appeal. Apparently self-hosted open source media library management is harmful.

Who knew open source software could be so subversive?"

I dont now if this is the correct community to post this. My computer shows signal to the screen when in the os (arch) but when I spam F2 or F12 to get into bios or boot selection it doesnt show anything to the screen. I have an optiplex 3020.

cross-posted from: https://lemmy.ml/post/31251815

I asked teachers to tell me how AI has changed how they teach.

The response from teachers and university professors was overwhelming. In my entire career, I’ve rarely gotten so many email responses to a single article, and I have never gotten so many thoughtful and comprehensive responses.

One thing is clear: teachers are not OK.

They describe trying to grade “hybrid essays half written by students and half written by robots,” trying to teach Spanish to kids who don’t know the meaning of the words they’re trying to teach them in English, and students who use AI in the middle of conversation. They describe spending hours grading papers that took their students seconds to generate: “I've been thinking more and more about how much time I am almost certainly spending grading and writing feedback for papers that were not even written by the student,” one teacher told me. “That sure feels like bullshit.”

AI companies claim their tools couldn't exist without training on copyrighted material. It turns out, they could and it just takes more work. To prove it, AI researchers trained a model on a dataset that uses only public domain and openly licensed material.

What makes it difficult is curating the data, but once the data has been curated once, in principle everyone can use it without having to go through the painful part. So the whole "we have to violate copyright and steal intellectual property" is (as everybody already knew) total BS.

The project implements sparse multiplication and fuses up/down projections in the MLP layers through low rank weight activations. Work is based on Deja Vu and Apple's LLM in a Flash.

This approach avoids loading and computing activations with feed forward layer weights whose outputs will eventually be zeroed out.

It's a lossless approach as these weights anyway do not contribute in the current token prediction. It does however, need the predictors to be accurate in clustering the weights.

The result? 5X faster MLP layer performance in transformers with 50% lesser memory consumption avoiding the sleeping nodes in every token prediction. For Llama 3.2, Feed forward layers accounted for 30% of total weights and forward pass computation resulting in 1.6-1.8x increase in throughput:

Sparse LLaMA 3.2 3B vs LLaMA 3.2 3B (on HuggingFace Implementation):

- Time to First Token (TTFT):  1.51× faster (1.209s → 0.803s)
- Output Generation Speed:     1.79× faster (0.7 → 1.2 tokens/sec)  
- Total Throughput:            1.78× faster (0.7 → 1.3 tokens/sec)
- Memory Usage:                26.4% reduction (6.125GB → 4.15GB)