Josh Armour

At Twitter, I believe we did a great job in interviewing. We interviewed in pairs (one talks, one takes notes), we used rubrics to ensure fairness and had lots of checks and balances. Everyone who did interviewing had to do training, shadow and then could participate. Although I wasn't involved in the final decision / offer side of it, I did interview a fair amount of it and even interviewed my new manager a couple of times. In all those interviews I never ONCE looked at a candidates school / college and certificates or qualifications. We always asked questions to figure out what they could DO, how they worked and their career in the sense of what they wanted and how they learnt from their journey. Evidence of emotional intelligence, learning and curiosity were key, beyond the technical skills. I still believe that was a great approach.

788

43 Comments

Pro tip: you can recognize the intellectually honest CrowdStrike outage commentators as those that present you now with pieces that they had originally written *before* the outage happened.

7

I've been developing, distributing and analyzing browser extensions for well over a decade, and consider myself pretty experienced in that area. The variety of extensions out there range from annoying adware, through silly gags, to productivity tools that could boost up your day to day performance. Even with the ever watching eye of Google over their store - malware finds a way. And it only becomes more and more sophisticated... That's why having trustworthy extension that can protect your browsing experience from such malicious (or even just plain annoying) instances - is very important. You can read more about it on our recently published ReasonLabs blog post: https://lnkd.in/dps3YBjr If you want to skip the reading and go directly to protecting yourself and your identity online, check out our Online Security solution: https://lnkd.in/dFS7ksPH Stay safe!

21

Today's Adobe security update has patched the in-the-wild zero-day PoC that EXPMON Public (https://pub.expmon.com) detected couple months ago (https://lnkd.in/ghQYYzXR). https://lnkd.in/gqxHHCfe Note: not want to cause confusion - this is not a working exploit, but a good PoC. Users are recommended to update their Adobe Reader ASAP, as we're going to point to the public sample with all the details soon in the next few days.

12

** There's New Development about this case ** We tested the (exactly the same) sample on the "patched" Adobe Reader version, it displayed additional dialogs, but if the user clicked/closed those dialogs, the app still crashed! Same UAF bug! We're contacting Adobe to find out what's going on and will update here. For now, we have to postpone the release of the details because the bug seems not properly fixed. [Update on Monday] Adobe has confirmed to us that the bug is not properly patched. "I can confirm our Acrobat product team has identified a secondary fix that is required to fully address this issue. We are actively reviewing and working to prioritize the fix in an upcoming patch. I will be sure to follow up with you once we have a clear release timeframe." Guess we have to hold our blog posts until the new patch becomes available to users. Sigh.. https://lnkd.in/gbBDuDHC #expmon #adobe #pdf #zeroday #0day #detection

8

Getting ready for my and Benjamin Perak’s reinvent talk by showcasing the power of Hooks and ECR. You can use Hooks to block the deployment of container images with critical vulnerabilities to CloudFormation. Pretty rad to just make your org safer with a few lines of code. See ya’ll next week! #reinvent #aws

53

2 Comments

https://lnkd.in/gtkbhUY7 I have been working on Multi-Agent systems for more than a year and I am still taken by surprise in their application/advancement. Here is an example that is a bit silly yet not really silly at all. It touches on the possibility of agents in simulation, which is where my current funding is targeted. Multi-Agent systems have already demonstrated: Simulating cooperative environments Coded new agents to accomplish emergent need Come up with new solutions to in distribution problems Multi-Agent systems have the potential to: Introduce Life-Long-Learning to generative models Use RL to learn to solve similar problems more effectively Come up with new solutions to out of distribution problems (learn to learn) and some other stuff I am not ready to share quite yet ;-)

3

1 Comment

Our little book from Accord Project maintainers is now on Amazon: "This book introduces computable contracts, also referred to as smart legal contracts. It emphasizes the integration of legal text with machine-readable data and executable logic as a solution to inefficiencies in contract management. It also explores the application of large language models (LLMs) and natural language processing (NLP) for computable contracts." https://lnkd.in/eZ8Fb_5F

43

4 Comments

A great engineering post by Matthew Castanien Upleveling at ZI is a long process and contains many wins like proper testing decommissioning via disable first to reduce blast radius in case something goes wrong (and you know usually it goes wrong 😅) Kudos for great tips how to make it safer on GCP

20

How do you #SRE? take this year’s Catchpoint survey and contribute to better tools and practices for reliability engineers everywhere

1

Happy Friday y’all!!! 🕺🕺 For those of who you couldn’t make it to talk Rudra Gupta and I did at OSFF in NYC, here is the video for the talk! 🎉 Don’t hesitate to reach out to either one of us if you’re interested in contributing to Common Cloud Controls (CCC) and Compliant Financial Infrastructure (CFI)! ☺️ #cloudsecurity #finos #commoncloudcontrols #compliantfinancialinfrastructure

28

This piece for Inc. gives tips on how to push back against the reflexively negative shut-down artists, highlighting how understanding failure can be another way to identify opportunities or to avoid another catastrophe altogether. Check it out, here.

1

Project update! My application tracking extension is now doing fuzzier matching and injecting new information into LinkedIn (and other sites) to keep me from doubling up on applications. Next up: local storage and bulk import/export operations...

12

2 Comments

LLMs are a fantasic tool, but know thier limits folks. Output value is inversely proportional to the level it needs to be trusted. If it _representing you_, write it yourself. If you really don't care if anyone else ever sees it, go nuts. But when is that ever true? For example: This is the output of a quick tool I wrote to summerize public record meeting transcripts. It does not represent any person. Never trust it as fact; http is deliberate. https://lnkd.in/eBddmpPq

3

My (not necessarily original) personal weekend thoughts on the Open Source Initiative's version 1.0 of the "Open Source" AI Definition. Special shout out to the Software Freedom Conservancy's position. Their write-up is much better than mine, but that's a very high bar I'd have to set for myself :-) https://lnkd.in/evRtz3BQ #OpenSource #OpenSourceAI #AI #LLMs #DigitalSovereignity

4