Posts

Showing posts from June, 2023

New GPT-4 powered Microsoft Shopping tools arrive on the new Bing and Edge #1197

Image
New GPT-4 powered Microsoft Shopping tools arrive on the new Bing and Edge #1197 Source  

Introducing Playground with Mixed Image Editing - AI Image Editor #1196

Image
Introducing Playground with Mixed Image Editing - AI Image Editor #1196 Source  

Salesforce Introduces XGen-7B, a new 7B LLM trained on 8K seq. length for 1.5T tokens #1195

Image
Salesforce Introduces XGen-7B, a new 7B LLM trained on 8K seq. length for 1.5T tokens #1195 Source

First drug discovered and designed with generative AI enters Phase II trials, with first patients dosed #1194

Image
First drug discovered and designed with generative AI enters Phase II trials, with first patients dosed #1194 Source  

AI Anchor Sana - Aaj Tak News India #1193

Image
AI Anchor Sana - Aaj Tak News India #1193 Source  

zeroscope_v2 - Open Source Text to Video Model #1192

Image
zeroscope_v2 - Open Source Text to Video Model #1192 Source  

Fast Segment Anything #1191

Image
Fast Segment Anything #1191 Source  

Stability AI - Stable Diffusion XL 0.9 (SDXL 0.9) #1190

Image
Stability AI - Stable Diffusion XL 0.9 (SDXL 0.9) #1190 Source  

Inflection-1, powering Pi - outperforming GPT-3.5, Llama and PALM-540B #1189

Image
Inflection-1, powering Pi - outperforming GPT-3.5, Llama and PALM-540B #1189 Source

Textbooks Are All You Need #1188

Image
Textbooks Are All You Need #1188 Source  

Marvel used AI to create opening intro for their new series: Secret Invasion #1187

Image
Marvel used AI to create opening intro for their new series: Secret Invasion #1187 Source   My Reddit Post Source   Source   Source   Source Source Source Source

MagicBrush: A Manually Annotated Dataset for Instruction-Guided Image Editing #1186

Image
MagicBrush: A Manually Annotated Dataset for Instruction-Guided Image Editing #1186 Source  

DeepMind: Introducing RoboCat, a new AI model designed to operate multiple robots #1185

Image
DeepMind: Introducing RoboCat, a new AI model designed to operate multiple robots #1185 Source  

OpenLLaMA 13B Released #1184

Image
OpenLLaMA 13B Released #1184 Source

MetaAI: Introducing Voicebox, a new breakthrough generative speech system #1183

Image
MetaAI: Introducing Voicebox, a new breakthrough generative speech system #1183 Source  

Progressively Optimized Local Radiance Fields for Robust View Synthesis (Better NeRFs) #1182

Image
Progressively Optimized Local Radiance Fields for Robust View Synthesis (Better NeRFs) #1182 Source  

Seeing the World through Your Eyes #1181

Image
Seeing the World through Your Eyes #1181 Source  

Google: How AI makes virtual clothing try-on more realistic #1180

Image
Google: How AI makes virtual try-on more realistic #1180 Source  

Augmenting Language Models with Long-Term Memory #1179

Image
Augmenting Language Models with Long-Term Memory #1179 Source  

Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation #1178

Image
Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation #1178 Source

OpenAI's ChatGPT June 13 Update - New Model versions, 16k context length for GPT-3.5, Lower API cost & more! #1177

Image
OpenAI's ChatGPT June 13 Update - New Model versions, 16k context length for GPT-3.5, Lower API cost & more! #1177 Source  

Introducing SlimPajama-627B: the largest extensively deduplicated, multi-corpora, open-source dataset for training large language models #1176

Image
Introducing SlimPajama-627B: the largest extensively deduplicated, multi-corpora, open-source dataset for training large language models #1176 Source  

Google's Generative AI support on Vertex AI is now generally available (PaLM 2 and more) #1175

Image
Google's Generative AI support on Vertex AI is now generally available (PaLM 2 and more) #1175 Source  

OpenLLaMA 7B trained on 1T tokens #1174

Image
OpenLLaMA 7B trained on 1T tokens #1174 Source

Great to see the world taking AI seriously, there's more than this... #1173

Image
Great to see the world taking AI seriously, there's more than this... #1173 Source Source Source Source  

Meta just released MusicGen, a simple and controllable model for music generation #1172

Image
Meta just released MusicGen, a simple and controllable model for music generation #1172 Source  

Tracking Everything Everywhere All at Once #1171

Image
Tracking Everything Everywhere All at Once #1171 Source  

Recognize Anything: A Strong Image Tagging Model #1170

Image
Recognize Anything: A Strong Image Tagging Model #1170 Source

Stability AI launches Uncrop, a new outpainting tool powered by Stable Diffusion XL #1169

Image
Stability AI launches Uncrop, a new outpainting tool powered by Stable Diffusion XL #1169 Source

AlphaDev, an AI system using reinforcement learning to discover enhanced computer science algorithms #1168

Image
AlphaDev, an AI system using reinforcement learning to discover enhanced computer science algorithms #1168 Source  

Text-to-3D Character From DAZ 3D #1167

Image
Text-to-3D Character From DAZ 3D #1167 Source  

Mustafa Suleyman on how AI will change jobs and improve future through intelligence #1166

Image
Mustafa Suleyman on how AI will change jobs and improve future through intelligence #1166 Source  

Meet LTM-1: LLM with 5,000,000 prompt tokens (coming in future) #1165

Image
Meet LTM-1: LLM with 5,000,000 prompt tokens (coming in future) #1165 Source  

VideoComposer: Compositional Video Synthesiswith Motion Controllability #1164

Image
VideoComposer: Compositional Video Synthesiswith Motion Controllability #1164 Source

INTELLIGENCE OFFICIALS SAY U.S. HAS RETRIEVED CRAFT OF NON-HUMAN ORIGIN (UFO) #1163

Image
INTELLIGENCE OFFICIALS SAY U.S. HAS RETRIEVED CRAFT OF NON-HUMAN ORIGIN (UFO) #1163 Source

Meta AI — Hiera is an extremely simple hierarchical vision transformer that's both more accurate than previous models + significantly faster at inference and during training #1162

Image
Meta AI — Hiera is an extremely simple hierarchical vision transformer that's both more accurate than previous models + significantly faster at inference and during training #1162 Source

Apple Vision Pro headset #1161

Image
Apple showed their Vision Pro headset for the first time, I am extremely impressed other than that it's not for gaming yet #1161 Source Source   Source   Source Source   Source   Source