Simon @spoltier@qoto.org

usually I'm very against the "write a bad draft and edit it to be good" approach (I feel like the "bad draft" just distracts me and I'd rather just delete it until the inspiration for a good version comes) but I guess right now it's kind of working for me

tried a new strategy with this terminal zine: sending a draft to beta readers that is genuinely just not very good (bad organization, basically just a collection of random facts) and getting some very early feedback. It feels bad but the feedback has actually been SO helpful, lots of positive comments telling me what the interesting bits are.

We are actively seeking a Security Engineer II, Application Security, to join our growing AI/ML Security team at Trail of Bits!

You'll get to work across the entire AI/ML stack - from examining LLM web applications and training pipelines to analyzing neural network architectures and MLOps environments. Our team conducts comprehensive security assessments of AI/ML systems, develops novel testing frameworks, and helps shape industry standards for AI/ML security.

If you would like to learn more about our work in the AI security space, check out some of our recent engagements and publications:

- AiLayer Labs https://github.com/trailofbits/publications/blob/master/reviews/2024-05-ailayerlabs-6079smartcontracts-securityreview.pdf

- Pickle File Attacks https://blog.trailofbits.com/2024/06/11/exploiting-ml-models-with-pickle-file-attacks-part-1/

- Audit Gradio 5 https://blog.trailofbits.com/2024/10/10/auditing-gradio-5-hugging-faces-ml-gui-framework/

Feel free to apply for the role below!
https://apply.workable.com/trailofbits/j/221A3565AB/

"We don't have to let the numbers play god with our inferences"

So I promised that @flourn0 and I were going to sit down & have a casual chat about that "ghost engineers" viral stuff we've been seeing in the news & our POV as scientists working on activity data in software development. So we did! And I posted my very first youtube video! We hope that this provides some help and support if you're a developer feeling overwhelmed by big claims like this!

https://youtu.be/e4gJxiDIFJ8?si=fS3U5p5XHZu72IGn

Systemd 257 released https://lwn.net/Articles/1001657/ #tech #linux

dtrace.conf is live and @bcantrill is mid-state of the union https://www.youtube.com/watch?v=KQN1t2qlhaw

Today is the day of dtrace.conf(24), our quadrennial DTrace conference! It's online and it's free -- join @ahl and me and a terrific slate of presenters starting at 9a Pacific!
https://bcantrill.dtrace.org/2024/12/05/dtrace.conf24/

@grimalkina I think it is in part a consequnce of the biases of social media. Lots more ADHD folks than dyslexic on written social media, dyslexics are on Instagram.

@grimalkina I'll second that after two decades in tech. In general, techies use of "neurodivergence" is for "cool" or "savant like" conditions. "Aspis" -- once outright fashionable but now dated. Synesthesia is one of the cool ones. PTSD and mood disorders are considered "mental health issues". "Super recognisers" -- cool, prosopagnosia not cool. Learning disabilities - ah, well, poor kids. Advocacy seems to be absent or very rare for anything considered to come with "less ability".

I've been curious about something for a long time. When people in tech say "neurodivergence", do you think about experiences like PTSD, mood disorders, learning disabilities, and other less commonly mentioned categories that also fall under the big tent of neurodivergence (e.g., not always just specifically things that there is a lot of online popular content about, like ADHD)?

Friday was my last day at ChipFlow! I'm very proud of what we were able to achieve, including major updates to Amaranth, CXXRTL, and the RTL debugger.

going forward I will continue to work, now independently, on Amaranth and other projects in the FOSS FPGA/ASIC tooling ecosystem, as well as on projects like ThunderScope

The impending #TikTok ban is a free speech ban that protects US Big Tech incumbents. It's in no way fundamentally different or better than other countries banning US websites or apps to protect their incumbents.

If the US wanted to protect private information from the Chinese government, it could pass comprehensive privacy legislation and regulate data brokers.

Instead, if enforced, we'll see app store censorship and ISP-level blocks impacting millions, with no material privacy benefit.

Fellow Signal users ​✨:

In case you use Signal but haven't discovered yet the wonderful feature that the "Note to Self" is, know that there is a wonderful feature called Note to Self !

It is amazingly useful!

HOW TO ❓

1. Just send a message to yourself in Signal to start using it! 🗨️​

2. You can use it with the automatic Disappearing Messages feature for more #privacy 🔥

3. You can Pin it at the top of your conversations for more convenience 📌

4. You can use it to transfer links, texts, and photos between your phone and computer in a safe end-to-end encrypted way! 🔒

5. You can change the background and bubble colors of your Note to Self thread to avoid getting mixed up and sending random personal notes to a random Signal contact 😉​

6. Magic!✨

Why Gelsinger was wrong for Intel
https://bcantrill.dtrace.org/2024/12/08/why-gelsinger-was-wrong-for-intel/

@dev 🥰
https://masto.ai/@cybeardjm/113610662636294807

So it’s beginning. White pelple can afford to ignore stuff like this. It makes me scared to go outside. https://me.dm/@anildash/113612287392541124