Nouscrafts

[Nouscrafts @nouscrafts]

@Grady_Booch Please! Please!

[tsotchke @tsotchke]

if you're a software engineer who's talented at C/C++, Python, distributed systems, GPGPU, and has considerable interests in physics-based and high performance computing looking to do some bleeding edge AI development please contact me

[Srishti @srishticodes]

Harvard just made degrees worth $200k obsolete by open-sourcing its Senior AI Engineer roadmap

Stop paying for bootcamps. Prof. Vijay Janapa Reddi just put the entire ML Systems (CS249r) curriculum on GitHub.

If you master these 6 pillars, you're ahead of 99% of the field:
> Architecture
> Data Pipelines
> Production
> MLOps
> Edge AI
> Privacy

This is the "Black Box" of Big Tech infrastructure, open-sourced.

Read. Learn. Bookmark.

Book - https://t.co/997GB0s1fl
GitHub Repo -https://t.co/y8pDr4lGFt

[hackaday @hackaday]

Neither Android nor iOS: DIY Smartphone Runs on ESP32! https://t.co/Q5ggaqFQOt

[China pulse 🇨🇳 @Eng_china5]

FUN BREAK:

Wi-Fi frequencies can map your body and movements even while you're inside your home! 👀

This Chinese project (WiFi-DensePose) proves that physical privacy is dwindling. Amazing technology, but frightening if it falls into the wrong hands.

[adafruit industries @adafruit]

Internet Archive free book: Printed Circuit Handbook

https://t.co/5dkSp5P3aT

[Mustafa @oprydai]

https://t.co/D1Bk777N9Q

[Adam Taylor @ATaylorFPGA]

Want to learn about how you can use the Versal AI engines for DSP.

I wrote a technical white paper here on it

https://t.co/cpfvZJe7SC

[Simplifying AI @simplifyinAI]

"I don't have a GPU" is officially dead 🤯

You can now run 70B model on a single 4GB GPU and it even scales up to the colossal Llama 3.1 405B on just 8GB of VRAM.

AirLLM uses "Layer-wise Inference." Instead of loading the whole model, it loads, computes, and flushes one layer at a time

→ No quantization needed by default
→ Supports Llama, Qwen, and Mistral
→ Works on Linux, Windows, and macOS

100% Open Source.

[Oliver Prompts @oliviscusAI]

Microsoft killed the GPU mafia 🤯

They finally open-sourced their 1-bit LLM inference framework called bitnet.cpp. It lets you run 100B parameter models on your local CPU without GPUs.

- 6.17x faster inference
- 82.2% less energy on CPUs

100% Open Source.

[Mustafa @oprydai]

https://t.co/imPc7ufYPB

[Nouscrafts @nouscrafts]

@AGDugin @grok , explain

[Mustafa @oprydai]

https://t.co/nVeby9otlp

[Nouscrafts @nouscrafts]

@simonateba @JeffDean @grok is it true?

[Nouscrafts @nouscrafts]

@br_resistencia @AGDugin Bicho, acorda, o problema do Brasil é justamente a elite quase toda analfabeta funcional. Potencial aonde com uma elite dessa?

[Nouscrafts @nouscrafts]

https://t.co/OpFRpv5ofX

[ΣTΞCH @SigmaTechRBLX]

Here’s the paper, I’ve read it so you don’t have to
https://t.co/CZ2sr4Cna7

The system requires at least 3 access points for Tx/Rx and assumes those access points remain stationary in known locations. This is enough for a prototype, but if you’re thinking someone could hack your system and place this software on your system, probably not

Even if you were compromised, you can easily foil such a system by nudging the access points around, having concrete or metallicized walls, or simply carrying a metal spoon to scatter the signals around. It’d still know you’re home, but not your dense pose.

An intruder would be better off just placing a camera inside your home if they really wanted to spy on you

The photo below are some failure cases from the paper

[Massimo @Rainmaker1973]

No cameras. No sensors.
Just Wi-Fi reading human movement.

Researchers at Carnegie Mellon University have developed a system that uses standard WiFi routers to detect human movement through walls—without any cameras.

By applying AI and DensePose model, the system turns WiFi signals into detailed maps of body posture in real time.

[Nouscrafts @nouscrafts]

@Grady_Booch @nvidia Huge fun!