Jade Sass

Solo and ghost dev | Software engineer | Deep Learning Machines

[Liu Liu @liuliu]

Happy to show what M5 Max can do with @drawthingsapp today at #wwdc. Come to check out at Mac & AI Track today and tomorrow. Happy to chat!

[Jade Sass @0x0jade]

@ICEWebDev @OfficialSamkayz Then you can still write puppeteer in Laravel

[Jade Sass @0x0jade]

@hackSultan AI doesn't have a mind of it's own, it follows a designed pattern for exploit, this pattern are then mixed with other future predictive patterns with different combinations making exploits seamless and well calculated.

[Yann LeCun @ylecun]

@eranshir @getnexar BADAS 2.0: new collision prediction system from Nexar.
It's based on V-JEPA 2!!
JEPA is going to save lives.

[Vaishnavi @_vmlops]

GOOGLE BUILT A SECRET WEAPON FOR FILE DETECTION

they ran it internally for years, gmail, drive, safe browsing, hundreds of billions of files every week

then they open sourced it

it's called magika and it exposes what files really are, not what they pretend to be

rename malware to "resume.pdf"? magika sees through it
disguise a script as an image? magika sees through it
any trick attackers use with file extensions? magika sees through all of it

ai trained on 100 million files. 200+ content types. 99% accuracy. 5ms per file

one command
`pip install magika`

the same tool protecting google's billion users is now protecting yours

https://t.co/Jr3LjmQobq

[Jade Sass @0x0jade]

@PovilasKorop You haven't met me, this is the exact way I'll approach such solution.

[Jade Sass @0x0jade]

Still the same Claude Code and Qwen3.5 27B running locally.

[Jade Sass @0x0jade]

Claude code + Qwen3.3 27B took around 2 minutes 34 seconds on 64GB RAM MacBook Max

[Jade Sass @0x0jade]

@abhijitwt Now his engineers gave us Claude's playbook for free...

[Tom Turney @no_stp_on_snek]

the original TurboQuant paper tested on A100 with models up to 8B.

6 days later, a bunch of strangers on the internet had it built and running on:
- Apple Silicon M1 through M5
- NVIDIA 3080 Ti through DGX Spark Blackwell
- AMD RX 6800 XT and 9070
- a 10-year-old Tesla P40
- an 8GB MacBook Air
- models from 3.8B to 70B across 6 architecture families
- 30+ independent testers

along the way we found new optimizations the paper didn't cover and failure modes it didn't test.

the fact that a loose group of people across the world can read a paper, build implementations from scratch, stress-test across hardware none of us could individually afford, and push the research further in under a week is genuinely one of the best things about this era. the tools and the community make it possible.

open source is something else.

[vixhaℓ @TheVixhal]

Most people with “AI/ML” in their bios don’t even know a real symmetric matrix always has real eigenvalues.

[Jade Sass @0x0jade]

@ellen_in_sf Literature review

[Mitko Vasilev @iotcoi]

I just implemented Google’s TurboQuant for vLLM.
My USB-charger-sized HP ZGX now fits 4,083,072 KV-cache tokens on GB10.

This may be the biggest open inference breakthrough of 2026 so far.

Training is the flex. Inference is the forever bill.

[Hugging Models @HuggingModels]

Meet a powerful new reasoning model that understands both text AND images. This 9B parameter model was distilled from top-tier reasoning architectures, bringing advanced chain-of-thought capabilities to a compact, efficient package. Perfect for developers who need multimodal smarts without massive compute.

[Lotto @LottoLabs]

Qwen 3.5 27b never degrades, never stops running, never has token limits, never refuses, never logs my prompts, never trains on my data, never sells my data, never runs up my credit card

[Taylor Otwell @taylorotwell]

17,000 requests per second on Laravel Cloud.

@dgarbs51_ breaks down exactly how he got there with k6 load testing - and what the results revealed.

https://t.co/80McjFYnz8

[NativePHP @nativephp]

⏰ Local Notifications Plugin
We’re about to make notifications… actually powerful.

Coming to NativePHP Mobile:
• recurring schedules (any frequency)
• action buttons → deep links
• Laravel tap events
• runtime CRUD

iOS + Android. One API.
Coming Soon 💪🔥

[Jade Sass @0x0jade]

@stanlee0nX Context memory overload...Your LLM is exhausting its capacity of context size.

[Jade Sass @0x0jade]

Me: attempting the OpenAI challenge parameter golf.