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
RESEARCH
-
Joscha Bach on Intelligence, Consciousness, and Reality
By
–
Here's my conversation with Joscha Bach (
@Plinz
), his 3rd time on the podcast. We talk about life, intelligence, consciousness, and the nature of reality. Joscha is one of the most brilliant and fascinating minds I know. This was mind-blowing and fun! https://
youtube.com/watch?v=e8qJsk
1j2zE
… -
CM3Leon: State-of-the-art text-to-image generation with less compute
By
–
CM3leon is a first-of-its-kind multimodal model that achieves state-of-the-art performance for text-to-image generation despite being trained with five times less compute than previous transformer-based methods. More details and examples https://
bit.ly/44I6t7E -
AI and nuclear fusion unite to shape future energy
By
–
Quand l'IA et la fusion nucléaire allument l'étincelle de l'avenir… Nous sommes à un moment charnière où l'IA et la fusion nucléaire s'unissent pour façonner notre futur énergétique. Un futur passionnant s'ouvre devant nous…
-
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
-
French Language Model Leaderboard Benchmark Platform Needed
By
–
ce serait super d'avoir un leaderboard comme https://
huggingface.co/spaces/Hugging
FaceH4/open_llm_leaderboard
… mais pour le français ! -
Untangling LLMs’ Training Data Sources and Ethics
By
–
From @benedictevans
' newsletter: the tangled web behind LLMs' training data -
Payment Intelligence: Unlocking Data Science Potential in AI
By
–
#PaymentIntelligence Unlocking Potential Data Science #AILabPage #VinsLens @KirkDBorne @SwissCognitive @alvinfoo @mclynd @SpirosMargaris @ipfconline1 @Fabriziobustama @FrRonconi @enricomolinari @ingliguori @schmarzo @asokan_telecom @SBUCloud @terence_mills
-
McKinsey AI Report Charts and Analysis Released
By
–
some charts from the McKinsey AI report released today
-
External Forces Impact Human Agency and Biological States
By
–
Thanks and really interesting. I think it’s critical to also recognize the external forces that can inspire or harm innate human / physiological states or biology to avoid what here feels a bit like “blame the human” to me. This logic can be leveraged to remove human agency.