"We then survey statistical lower bounds that, we argue, constitute a compelling case against the possibility of designing high-accuracy LAIMs with strong security guarantees."
On the Impossible Safety of Large AI Models
https://arxiv.org/abs/2209.15259
#generativeAI #chatBots #LLMs #safety #personalSafety #robustness #genAI #accuracy
Large AI Models (LAIMs), of which large language models are the most prominent recent example, showcase some impressive performance. However they have been empirically found to pose serious security issues. This paper systematizes our knowledge about the fundamental impossibility of building arbitrarily accurate and secure machine learning models. More precisely, we identify key challenging features of many of today's machine learning settings. Namely, high accuracy seems to require memorizing large training datasets, which are often user-generated and highly heterogeneous, with both sensitive information and fake users. We then survey statistical lower bounds that, we argue, constitute a compelling case against the possibility of designing high-accuracy LAIMs with strong security guarantees.
This Is What a Personal Surveillance System Actually Looks Like
You stop thinking of it as surveillance. It becomes “the system.” Just part of how things run.
OSINT Isn’t About Skill Anymore. It’s About Systems
There’s a moment, if you spend enough time in this space, where your attention changes. You stop looking at individual data points as things to extract. You start seeing them as inputs to a larger structure.https://cha1nc0der.wordpress.com/2026/03/21/osint-isnt-about-skill-anymore-its-about-systems/
His last name was spelled backwards on his Facebook Dating profile. Her mom figured out why from the parking lot just in time.
“There aren’t many non-fictions, self-help books that I would consider “page-turners,” but that’s exactly what Dale Yeager has written." Marjory Abrams, former Editor in Chief, Bottom Line Personal https://www.amazon.com/Profiling-Your-Life-Criminal-Dangerous/dp/B0D2Z7BZXN/ref=cm_cr_arp_d_product_top?ie=UTF8
profilingyourlife #personalsafety #selfdefense #forensicpsychology #protectyourfamily #criminalprofiling #truecrime #daleyeager
Collective adaptive goes virtuous circle:
- sharing information,
- creating opportunities for people to share the "messy details" of their experience,
- inquiring into team member’s ways "to expand what goes well",
- drawing from the capacity across the work collective.
https://www.infoq.com/articles/adapt-surprises-software-reliant-businesses/
#collectives #teamWork #workCulture #facilitation #management #IT #DevOps #resilience #engineering #fallBack #recovery #psychologicalSafety #personalSafety #safety #feedback #joy #curiosity
This article explores understanding what makes incidents so rare (when and how they do not happen) and so minor (over how much worse they can be) and deliberately enhancing what makes that possible.
“There aren’t many non-fictions, self-help books that I would consider “page-turners,” but that’s exactly what Dale Yeager has written." Marjory Abrams, former Editor in Chief, Bottom Line Personal https://www.amazon.com/Profiling-Your-Life-Criminal-Dangerous/dp/B0D2Z7BZXN/ref=cm_cr_arp_d_product_top?ie=UTF8
profilingyourlife #personalsafety #selfdefense #forensicpsychology #protectyourfamily #criminalprofiling #truecrime #daleyeager
Eric Maugendre on collectives
Chaincoder
FurballsNHairballs
Front Page
Dale Yeager