We’re gonna be stuck on SDXL, aren’t we?
@hardmaru
-
Reusing Old NeurIPS Paper Template After Three Years
By
–
The last time I published a NeurIPS paper was back in 2021, so I just reused my old template from 3 years ago. I’m not sure we will submit to NeurIPS, since I heard that the process it is quite difficult these days 🙂
-
Celebrating Open-Source Robotics Growth and Community Development
By
–
Congrats! The world needs more open-source robotics!
-
Alternative Timeline: Hebbian Learning Dominance in AI
By
–
Imagine the alternative timeline if Hebbian Learning took over…
-
1-Bit Neural Networks: Not New, Earlier Research from 2016
By
–
1-bit neural nets are not new. Earlier papers from 2016: Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1 https://
arxiv.org/abs/1602.02830 Ternary Neural Networks for Resource-Efficient AI Applications https://
arxiv.org/abs/1609.00222 -
Are Video Generation Models World Simulators?
By
–
Are Generation Models World Simulators? https://
artificialcognition.net/posts/video-ge
neration-world-simulators/
… -
Question about Clipdrop app sale
By
–
wait, but didn’t they just sell @clipdropapp? https://t.co/1UwKFVsfId
— hardmaru (@hardmaru) 29 février 2024wait, but didn’t they just sell @clipdropapp
? -
Action-Conditioned 2D Models Show Promise for Future Progress
By
–
I think even with action-conditioned 2D, we will get a long way
-
Mental World Models vs Video Generation: Survival-Optimized Cognition
By
–
Our own mental world models are probably not as capable as video generation models, as in we can’t ‘generate’ a stream of high resolution detailed images at 30fps. Instead, our mental world models are highly optimized to facilitate our survival as an individual and as a species.
-
AI Tools Users Bear Responsibility for Outputs and Actions
By
–
Just like any tool used by humans in the past, individuals or institutions using AI tools should be responsible for the outputs or actions produced by such tools.