LATEST AI Advances: Dreambooth, Midjourney V4, Photorealistic Text to Image Model & Google Imagen

TL;DR

This video discusses the latest AI news, including the advancements in Mid Journey V4 surpassing Dolly 2, the impressive results of applying Mid Journey V4 to Stable Diffusion using Google's Dream Booth, the ethical considerations of training models on other AI's images, Lexica's coherent and high-resolution model, and updates in Playground AI and Runway ML.

Transcript

hello viewers and welcome back to another video today we're going to be doing some AI news this is a new series I've sort of started I have a whole playlist on the YouTube channel for it we did one last week and I'm just rounding up all of the latest and greatest AI news that I've found through especially my Discord server but also on Twitter and a... Read More

Key Insights

• ❓ Mid Journey V4 surpasses Dolly 2 in coherency and artistic ability, demonstrating its potential for generating realistic and imaginative images.
• ✋ Lexica's model shows impressive coherence and high resolution, indicating the progress in AI text-to-image generation.
• 👶 Playground AI and Runway ML have introduced new features, including infinite image and image-to-image generation, expanding the possibilities of AI creativity and video editing.

Questions & Answers

Q: How does Mid Journey V4 compare to Dolly 2 in terms of coherency and artistic ability?

Mid Journey V4 has been deemed superior to Dolly 2 in terms of coherency and artistic ability, as it can generate realistic images combining concepts like a Darth Vader toilet or a whale in bed.

Q: Is it ethically justified to train models using Mid Journey's images without permission?

The video suggests that it is ethical as the purpose of AI models is to learn and improve by drawing from other models' creations.

Q: What features does Lexica's model offer?

Lexica's model demonstrates high coherence and resolution, as seen in its ability to generate realistic images. It shows potential with its improved text-to-image generation compared to other models.

Q: What changes have occurred in Playground AI's pricing model for Dolly 2?

Playground AI now charges separately for Dolly 2, with a $10 fee as an add-on, while previously it was included in the Pro model priced at $30. This change has received mixed reactions.

Summary & Key Takeaways

• Mid Journey V4 has surpassed Dolly 2 in terms of coherency and artistic ability, as demonstrated by its ability to generate images like a Darth Vader toilet or a whale in bed.

• Lexica, a website similar to others in generating AI text-to-image models, is working on its own model that shows great coherence and high resolution.

• Runway ML has introduced new features like infinite image and image-to-image generation, while Playground AI has changed its pricing model for using Dolly 2.

• Google's AI Test Kitchen app, currently in a waitlist phase, showcases Google's Lambda model and will soon include Google Imaging.