This looks amazing. Virtual psychedelics! Happy to Beta test if you need one. 😉
GENERATIVE AI
-
Microsoft Copilot DALL-E 3 Improves Image Segmentation Capabilities
By
–
Has anyone noticed that the Microsoft Copilot version of Dalle-3 has gotten much better? You can segment any section of the image now. It seems we can only blur what's behind the segment or add more color to the segmented object but it's definitely getting better.
— Matt Wolfe (@mreflow) 8 février 2024
The… pic.twitter.com/MGTCQU2GvQHas anyone noticed that the Microsoft Copilot version of Dalle-3 has gotten much better? You can segment any section of the image now. It seems we can only blur what's behind the segment or add more color to the segmented object but it's definitely getting better. The
-
MLX Special Tokens Dataset Training Implementation
By
–
One thing I realized you did differently Mike, you added the [INST] Instruction [/INST] in the text string of your dataset. I was under the impression that MLX automatically adds the special token based on the original model, so you don't have to do that. But I may be
-
LoRA vs Quantization: Optimizing AI Model Training Resources
By
–
Yes! I could have done Lora as well but I opted for the quantized version to reduce resources. If I had more than 64gb of RAM I feel Lora would have been fine.
-
Microsoft pivots Copilot strategy from search to AI assistant
By
–
Love that Microsoft is going all in with Copilot, and positioning it more as an AI assistant rather than a search tool (Bing)
-
ChatGPT Performance Decline Leak Impacts AI Market
By
–
Ah thanks for sharing! We included this in the newsletter for tomorrow but forgot to add it to the thread. The leak adds up, given the recent tanking of ChatGPTs performance
-
Outtakes AI Generates Immersive Worlds for Apple Vision Pro
By
–
Outtakes AI posted a demo of its app showcasing the ability to generate immersive worlds on the Vision Pro.
— Rowan Cheung (@rowancheung) 8 février 2024
Apple's headsets + AI are going to enable some truly insane experiences. The Holodeck is becoming a reality.pic.twitter.com/iJ9eKNSzMCOuttakes AI posted a demo of its app showcasing the ability to generate immersive worlds on the Vision Pro. Apple's headsets + AI are going to enable some truly insane experiences. The Holodeck is becoming a reality.
-
Stability AI Updates Stable Video Diffusion Model With Enhanced Motion
By
–
Stability AI released the first update to its Stable Diffusion model, with enhanced motion and consistency in generations. Will 2024 be the year of open-source AI?
https://
x.com/i/status/17538
58278868545833/video/1
… -
MGIE: AI Image Editing via Natural Language Prompts
By
–
Apple and UC Santa Barbara researchers introduced MGIE, an open-source AI system for editing images via natural language prompts.
— Rowan Cheung (@rowancheung) 8 février 2024
Image editing is about to get even easier: pic.twitter.com/0Txmnvm6NjApple and UC Santa Barbara researchers introduced MGIE, an open-source AI system for editing images via natural language prompts. Image editing is about to get even easier:
-
YouTube Bets on Generative AI to Empower Creators in 2024
By
–
YouTube is betting big on AI in 2024. Generative AI was listed at the top of CEO Neal Mohan's latest blog, highlighting his vision for empowering creators. The more popular platforms that embrace AI, the easier it becomes to shift public sentiment.