"You should get paid twice at when you're at work. You should get paid your check and then you should get paid through education."
@cheferick is absolutely right, whether it's in a restaurant like his or in other fields such as cybersecurity, this idea is so important
After honing his skills for more than two decades, @cheferick opened up his own fine-dining restaurant in Chicago called “Virtue” that’s serving up southern-based food and his “great grandmother’s hospitality.” #TheDish
@ImposeCost I also reject that it immediately means "do better." Sometimes you're lose. I'm team "learn from this."
You might find you made mistakes, so you do need to do better. You might find that you did everything right, but the other side has a plan as well
GRT2 at DEF CON is for everyone. Check out the blogs at https://t.co/6UAyeksKMR and to learn about the event and prepare to generate LLM reports. We’re also hosting a workshop at DEF CON 32 to enable anyone to try to capture-the-flags and win a cash prize ($50 - $200).
Mark your calendar for Keynote Speaker Nikki Pope’s talk in the AI Village room at DEF CON 32 at 10 am on Aug 10. Make sure you show up early, as seats will go fast!
Tomorrow from 4-8 PM Eastern on https://t.co/F2syuFJbqP - 4 hours of chaos for the kids, with @Darkness429, @AimsTV, and @BasicallyIDoWrk.
Help show the world, and the kids at @StJude, that gaming is a force for good. ❤️💜
Very excited that I will be presenting my ML security work (@trailofbits) at HOPE XV (@hopeconf)! I’ll be talking all about incubated ML exploits. Catch me tomorrow at 4 PM!
https://t.co/oGTmMNVX2t
The BSidesNoVA 2024 CFP closes TOMORROW NIGHT!! Get your submission in as soon as possible so we can get it to the review board!
Good luck to all the submitters so far <3
For model devs releasing LLMs in the open through @huggingface, it's currently impossible to protect those against malicious finetuning. But what if there was a way to "immunize" them?
In this work led by @domenic_rosati, we do exactly that! 1/5
In #BHUSA trainings “Black Hat Machine Learning” this course will provide students with a realistic environment and methodology to explore the unique risks presented by the use of ML in today's environments.
Register now >> https://t.co/DOSPjXjykP
This. The problem with cybersecurity is that almost everything that is taught about it is based on high level abstractions generated by people who never found a single software vulnerability in their lives.
This is a problem because cyber offense works on a concrete level, and with concrete vulnerabilities. ALWAYS! When a hacker attacks an organization, he doesn't attack some abstract Asset X or Perimeter Z, he attacks a Specific Vulnerability in Specific Code of Specific Software with an Extremely Specific Exploit (although code flavours and techniques may vary). When you don't have a practical grasp of this low level, your cyber defense is guaranteed to look a bit like this to us:
(I'm processing interest responses from my upcoming Cybersecurity course - with a little help from ChatGPT, of course, but this one caught my attention)