There was a reason the Japanese never invaded the mainland US in 1941, that reason has 100x’d since.
SAFETY
-
Comprehensive Review of The Atomic Human and Trustworthy AI Blueprint
By
–
Very comprehensive and generous review of The Atomic Human by @glawton Meet The Atomic Human – a draft blueprint for trustworthy AI
-
Three Eras of Artillery: From Indiscriminate to AI-Controlled
By
–
The three eras of artillery:
Past: indiscriminate
Present: GPS-guided
Future: AI-controlled -
Neural Networks Still Fail at Basic Tasks Despite Progress
By
–
In 1969, Minsky & Papert showed that two-layered neural nets couldn’t learn to determine whether a figure was connected or not. Fast forward to 2024, with vastly more data, layers & compute, and neural nets (as we now know how to make them) still fail at some really basic stuff:
-
Reducing Contextual Hallucinations in Large Language Models
By
–
5/ Contextual Hallucinations Mitigation in LLMs – proposes a new method that detects and significantly reduces contextual hallucinations in LLMs (e.g., reduces by 10% in the XSum summarization task)…
-
AI-Generated Videos: Growing Risk of Deepfakes and Manipulation
By
–
Et c'est à partir de maintenant qu'il va falloir faire attention aux videos que l'on va voir, celle du tireur compris #AI
-
AI Image Generation Speed Versus Real Imagery Authenticity Today
By
–
wrapping my head around how wild it is that we live in a moment now where you can just conjure up any image in like 2 seconds and yet some of the most unbelievable images i've seen showed up here today and they were real.
-
Deep Learning: Benefits, Challenges, and the Black Box Problem
By
–
5/5 Benefits & Challenges: While deep learning enhances accuracy, automation, and adaptability, it also faces challenges like high data requirements, computational cost, and the "black box" nature of its decisions. #AIChallenges #TechBenefits Follow @GiulianoLiguori for
-
LLMs Face Hard Problems of Messy Reality Like Robotics
By
–
Enjoyed @AlexIrpan
’s “The Tragedies of Reality Are Coming for You” The hard problems of robotics are not unique to robotics. In fact, as LLMs start to actually work and become widespread, they encounter more problems from the “messy reality” of the world.