#AI industry and researchers sign statement warning of ‘extinction’ risk https://
cnn.it/3C0kqBB #ethics #FutureofWork
SAFETY
-
AI Industry Leaders Warn of Extinction Risk from Advanced Systems
By
–
-
Science fiction inspires future through AI and faith storytelling
By
–
Sci-fi is important because it inspires the future. @minafahmi
's piece in @every today is an example of the best of that ethos. It's a story about science and faith in a world governed by a superintelligent AI. You should read it: -
Overlooked AI Risks: Data Bias in Healthcare, Finance, Justice
By
–
Townsend shares there are areas of #AI risk that can be overlooked now with hype surrounding GAI: "Things like biases in data that produces results that impact people today in healthcare settings, in financial services settings, in justice and public safety settings…" @fox5ny
-
OpenAI Closes AI Text Detector Due to Low Accuracy
By
–
"Classifier", OpenAI's AI-generated text detector, shut down due to its low accuracy rate https://actuia.com/actualite/classifier-le-detecteur-de-texte-genere-par-lia-dopenai-ferme-en-raison-de-son-faible-taux-de-precision/
… #AI #artificialintelligence -
RLHF Limitations and Open Problems in AI Alignment
By
–
RLHF(Reinforcement learning from human feedback) is the common technique used for aligning AI systems(LLMs in particular) with human goals. RLHF works but as with every thing in life, it has flaws too. The paper "Open Problems and Fundamental Limitations of Reinforcement
-
AI Systems Becoming Incomprehensible: The Control Problem
By
–
The Coming Wave is producing new inventions that are increasingly beyond our ability to comprehend or control… Previously creators could explain how something worked, why it did what it did, even if this required vast detail. That’s increasingly no longer true. Many
-
Untangling LLMs’ Training Data Sources and Ethics
By
–
From @benedictevans
' newsletter: the tangled web behind LLMs' training data -
PhotoGuard Protects Images Against AI-Driven Deepfakes
By
–
PhotoGuard helps combat the misuse of AI-driven image manipulation, including deepfakes. It uses invisible perturbations to disrupt AI’s ability to alter images, offering protection from fraudulent or malicious use: https://
bit.ly/3Ozt0yc -
IP Asymmetry: Valuing AI Corporate Rights Over Human Agency
By
–
Not saying that’s your aim – just that this type of logic can be pervasive and has deep irony / can cause harm when IP for companies creating AI is recognized and valued and “IP” of human data and agency is not. Also – people are not inherently broken. Free will is also a thing.