Soon, nearly all text, image, video, and audio content will be 100% (or partially) generated by AI. We won’t be able to tell what’s real from what’s fake. The fact is, the question is no longer whether this is good or bad. It’s about how we will adapt and evolve in this society.
@dfintelligence
-
ChatGPT Already Uses RLHF
By
–
— "Training that would involve interactions with players and use the RLHF (Reinforcement Learning from Human Feedback) technique." ChatGPT already employs these methods to communicate. What would have been more accurate would have been to discuss "fine-tuning" for chess-related tasks.
— -
AI and MRI to reconstruct dream images
By
–
AI image generation models could allow you to see your dreams in images. Japanese researchers showed images to people and then reconstructed the images using brain MRI data. Here are the results + source in the second tweet.
-
Go see the new video by DEFAKATOR_Off on ChatGPT
By
–
Go see the new video by @DEFAKATOR_Off on ChatGPT!
-
Mobile hardware and AI: Snapdragon chips for edge computing on Qualcomm stand
By
–
Well, mobile isn't just screens; it's also a lot of less visible hardware at the @Qualcomm booth. The tone is to show the famous Snapdragon chips and the latest generations for AI. Edge computing on mobile.
-
ControlNet: new AI to guide Stable Diffusion with input image
By
–
ControlNet is the latest innovation in AI image generation. It will 'guide' Stable Diffusion to draw more inspiration from the input image. For example, it can be used to recreate faces… Alright, little thread, let's start with a Real version of Zelda
-
Where did the 1.5 billion euros for AI in France go?
By
–
In France, 1.5 billion euros were injected to make France a global leader in AI. Does anyone know where this money went? It costs 5 million euros to train a model like ChatGPT, or 10 million total with data and development. So where is the French ChatGPT, please?
-
LIVE Twitch: practicing AI with CUDA, ControlNet and GPT3
By
–
LIVE TWITCH https:// twitch.tv/defendintelligence … We are going to practice Artificial Intelligence. It's CUDA against me. And we play around with ControlNet and GPT3.
-
Complex AI setup with CUDA, VS Studio, PyTorch, dependencies
By
–
AI is cool and all until you have to install CUDA, VS Studio for C++ support, PyTorch, and manage dependency versions with the 154 other libs.
-
3D face of Lincoln generated by AI with GANs
By
–
Le visage de Abraham Lincoln en 3D à partir d'un modèle d'#IA ?
— Defend Intelligence (Anis Ayari) (@DFintelligence) 23 février 2023
Construire votre visage en 3D à partir d'un modèle d'ia c'est possible grâce à des méthodes développé par nvidia et l'université de Stanford en utilisant des GANs 3D.
Abonnez vous pour suivre l'actu de l'#IA ⬇️ pic.twitter.com/BVqDqLLt6zThe 3D face of Abraham Lincoln from an #AI model? Building your face in 3D from an AI model is possible thanks to methods developed by Nvidia and Stanford University using 3D GANs. Subscribe to follow the #AI news.