Harmless supernova fallacy, "bounded therefore harmless". https://
arbital.com/p/harmless_sup
ernova/
…
SAFETY
-
Harmless Supernova Fallacy: Bounded Therefore Harmless
By
–
-
Understanding Giant Tensors and Writing Goodhart-Proof Objectives
By
–
It doesn't need to be representing new physics, for us to have trouble understanding what the giant inscrutable tensors mean — or writing airtight Goodhart-proof objectives that operate over them, a much higher requirement of proficiency than current interpretability efforts.
-
AI Objectives: Learned vs Human-Coded Implementation Types
By
–
It's also described in my "Creating Friendly AI" from 2001. What sort of objectives do *you* have in mind? Learned? Human-coded? What's their type signature?
-
MLflow-Giskard Integration for LLM Testing and Vulnerability Detection
By
–
The MLflow-Giskard integration offers a great solution for testing and validating LLM outputs. MLflow's evaluation API + Giskard's automatic vulnerability detection for LLMs is game-changing. See it in action https://
dbricks.co/4bEoPKl -
Chollet’s Debate Participation and Availability Questions
By
–
Maybe it turns out Chollet doesn't like debates. Has he done any others? When did he sign up to be in debates and not just (end the world via) building AI frameworks? I'm up for it if he is, but don't want to be the guy claiming he's got to provide that service for free.
-
Unconstrained Solution Search in Natural Selection and SGD
By
–
Solving some outer reward or loss via a process that looks around for solutions and isn't much constrained or much legible in which solutions it finds. Natural selection and stochastic gradient descent are both examples.
-
Hugging Face Security Breach Exposes Space Secrets
By
–
https://
huggingface.co/blog/space-sec
rets-disclosure
… -
Google’s AI goes crazy: explanation
By
–
L’IA de Google pète les plombs ? 🤯
— Defend Intelligence (Anis Ayari) (@DFintelligence) 31 mai 2024
Je t’explique ! ⬇️ pic.twitter.com/aQf16zYbvTGoogle's AI goes crazy? I'll explain!
-
LHC Black Holes to AGI: History Repeating Fear Cycles
By
–
this same LHC which was feared to potentially create black holes that would destroy the earth in mainstream news 🙂 –history keep repeating itself. Yesterday fearing black-holes from uncontrollable particules-collider, today fearing terminator-like risk from uncontrollable AGI
-
AI Insurance Sector Faces New Technology Risks
By
–
As generative AI becomes more prevalent, the insurance sector faces new risks, especially in technology. Addressing these emerging challenges is crucial for ensuring safe and effective AI adoption across industries.