
Web, Mobile, IoT: all JS things since 00's - opinions are my own ™ Principal SW Engineer @anacondainc working on OSS stack 🦄
Milan, Italy
Joined May 2009
- Tweets44.7K
- Following1.3K
- Followers10.2K
- Likes15.7K
Replying to @alexellisuk
@alexellisuk how to DM you? Meantime, once switchless-nccl is installed, both Nvidia sync and Sparkrun are incapable of making it work, any quick hint/pointer ??? Your fabric-setup won’t help neither
Replying to @satya164
Yeah, I meant something broader and standard than your Claude rule, but good for you that works
what's easier?
an AI on your GitHub account behalf is not you, it's an AI.
if a bot disagrees/brings up slopy things, the better model explains what is it that previous review missed and why that's not a concern.
we'll all have data around which model does good reviews/PRs! 🥳
is this idea silly?
automated/AI PRs should explicitly state the model/variant used to file such PR
automated CI/bots should explicitly state the model/variant used to approve/reject
automated amends/follow ups should explicitly state the model/variant used
all will be easier
Replying to @Nosredna
yeah, I guess my point is that nim, or any other PL not automatically embedded in this AI wave forward, is not a skill anymore, or not a skill at all; maybe Zig but if AI won't ever produce any of those PLs out there, they are effectively dead because not needed ... sad yet true.
apparently agents work like everything is:
* Python for tests/proofs
* JS/TS for Web/BE
* Rust/Go once Python validates &/|| JS works
* Swift on iOS/Apple
* Java/Kotlin on Android
* Qt/Gtk on Linux
* whatever MS uses 🤷
do we need any other programming language at this point?
Andrea Giammarchi 🍥 retweeted
So basically, today PHP got two critical bugs that let anyone run code on your server without even logging in, which is bad, and then the details leaked before the patch was even out, which is worse, and it works on WordPress and MediaWiki and Magento and Joomla and Laravel, and Censys says like 1.5 million servers are running a version with no fix, so the NSA was like "take your system offline," and the FBI was like "take your system offline," and CISA and the UK and CERT-EU were all like "take your system offline," and then the Prime Minister of Japan, the actual Prime Minister, posted at 5:55 in the morning telling 48.8 million people to take their system offline, so everyone took their system offline, like the Dutch cybersecurity agency took its own website offline, and the Dutch parliament went offline, and the Port of Rotterdam went offline, and Finland and France and Denmark started pulling government sites, and Wikipedia went down, and Archive.org went down, and Asana and GOG and Something Awful went down, and the University of Vienna's webmail went down for some reason, and that's just PHP. Because GitLab also dropped a CVSS 10.0, which is the highest score they give out, for a bug that's been there since 2014, so that's every self-hosted GitLab for twelve years, and then LiteLLM 1.107.0 got hijacked and shipped a credential stealer and a remote access trojan in every single wheel and the Docker images, including latest, and you didn't even have to import it, just installing it was enough, and it's the second time this year, and then Railway paused every deployment on the entire platform because someone got into the environment variables, so now everyone has to rotate every secret they've ever stored there, and TiDB Cloud said someone was walking around inside with stolen credentials for basically a month and an extortion group says they took 10.6 terabytes, and Zeabur got breached for the second time in two months and is just shutting down, like the whole company, and Node.js patched remote code execution and a TLS bypass, and CISA says Plex is being actively exploited, and someone took over 300 Tor exit relays to steal SSH passwords, and SafePay ransomware has hit over 600 critical infrastructure orgs, and India's biggest bank filed a data loss incident, and Japan has been in an official state of emergency in cyberspace since October 9, which honestly explains the 5:55 AM post. Oh, and eMudhra, a certificate authority that's issued over 60 million certificates, whose entire job is proving websites are who they say they are, was supposed to be using private keys that are random numbers over a hundred digits long, and one of them was 4. Just 4. And the others were like 66 and 237. Anyway, it's Wednesday.
Andrea Giammarchi 🍥 retweeted
This week, Drama TensorFold clones voices!
Give it a voice you have permission to use, and it speaks in that voice on your own GPU.
🔊 Meet Tomorrow, my assistant, running on it. Sound on!
github.com/plotarmordev/dram…
Replying to @matteocollina
Has anyone else answered DeepSeek already?
Andrea Giammarchi 🍥 retweeted

Mia@MiaAI_labOct 20
West Asia
West AsiaConnected via North America App StoreAccount-level information, not a live location or per-post device.
DeepSeek V4.1 Flash 1017 on 2x DGX Sparks
Frontier at home. No limits. 100% Private.
github.com/MiaAI-Lab/DeepSee…
Ray Fernando@RayFernando1337Oct 20United States
United StatesConnected via United States App StoreAccount-level information, not a live location or per-post device.
How is DeepSeek v4.1 Flash this fast on 2 DGX Sparks??? This is a national security risk.
Andrea Giammarchi 🍥 retweeted
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
🔗 github.com/deepseek-ai/DeepI…
1/4
Andrea Giammarchi 🍥 retweeted

Andrew Curran@AndrewCurran_Oct 19
Canada
CanadaConnected via Canada Android AppAccount-level information, not a live location or per-post device.
The Department of Homeland Security is requiring all civilian federal agencies to remove DeepSeek products and 'any third-party services developed using DeepSeek implementations,' and to stop buying them.
Replying to @srcbuild
“Not quite my tempo” … I have massive plan/goal and agents are working incrementally to adjust compaction, what are you talking about? 🤔
Andrea Giammarchi 🍥 retweeted

Mia@MiaAI_labOct 19
West Asia
West AsiaConnected via North America App StoreAccount-level information, not a live location or per-post device.
Here's how 4 concurrent streams look on DeepSeek V4.1 Flash 1017 running on 2x DGX Sparks.
Mia@MiaAI_labOct 19West Asia
West AsiaConnected via North America App StoreAccount-level information, not a live location or per-post device.
Run DeepSeek V4.1 Flash 1017 on 2x DGX Sparks ✨
- 1M total KV pool, 512K context
- 75 tok/s on prose, single stream
- 130 tok/s on prose, 4 concurrent streams
- Prefill 2000-2175 tok/s
- EXL3-EAQv1, 2.9 bpw
The quant in this recipe got about the same calibration as my GLM 5.3 Flash, with some extras. Feedback would be appreciated!
Special thanks to @bertholomusai and all the contributors for the engine work, and to @ashxhart for TensorFold.
Get it here:
github.com/MiaAI-Lab/DeepSee…
about the "missing" bit, compaction and STATUS.md theoretically solve that but it's nowhere visually easy to follow on a timeline based board ... anyone that already solved this problem out there? thanks in advance for pointers 👋
underrated skill in this AI era: the scrum master!
* definition of done (any AI loop)
* split bigger tasks into incremental milestones (context compaction when needed)
* story points per each task (energy consumption/deliverable ETA)
missing: an automated AI driven kanban board




