🚀 Midjourney Office Hours Highlights: September 18th

In the latest Midjourney Office Hours on September 18th, 2024, the team shared a bunch of exciting updates and projects they're working on. Check out the comprehensive summary of what was shared. If you prefer to just get the gist of it, jump down to the TL;DR; section.

Extended Recap

Here's a detailed recap of what was shared:

Personalization Features 🎨

Midjourney is set to launch a basic personalization page soon, aiming to simplify user access to personalization features. This initiative is designed to increase usage by making personalization one of the platform's standout features. Looking ahead, advanced personalization options are planned, including multiple profiles for style and personality-based customization. However, the team faces challenges with the fragile and complex algorithm and the difficulty of measuring effectiveness across a diverse user base. The ongoing debate centers on balancing ease of use with advanced features.

Model Updates 🤖

The upcoming Midjourney Version 7 promises significant improvements in knowledge integration, aesthetics, and prompt understanding. The new version will feature a simplified architecture to support more features, enhanced upscalers, and better character/style reference capabilities. It will also cater to 3D and video, in addition to 2D images. However, the training for V7 has encountered some issues, potentially delaying its release by 1-2 weeks. The focus has shifted from V6.2 to V7 to ensure better model performance.

Explore Page Improvements 🔍

The explore page is undergoing significant changes to improve user experience. The algorithm is being updated to transition to a "Curate Community Front Page" task for better image selection. The feed will be reorganized, removing the "random" and "hot" sections and introducing "new" and "top" sections. The "new" section will be automatically curated, while the "top" section will feature the best content. UI enhancements, including masonry layout changes, are also in the works to better display various aspect ratios.

Editor Development 🖌️

A standalone editor is being developed, allowing users to edit images separately from the creation flow. This editor will incorporate depth references and retexturing features. Future enhancements include session-based editing to avoid clutter from multiple iterations, advanced editing tools, and potential drag-and-drop functionality for ease of use.

3D and Video Models 🎥

Midjourney is scaling up data collection for improved 3D model training and integrating video models based on the V6 or V7 foundation. The team is working on simple and cost-effective video generation, with a long-term vision towards real-time open-world simulations. Future integrations may include VR applications for interactive 3D experiences and animated 3D scenes.

Psychometrics and Demographics 🧠

The team is exploring the links between aesthetics, personality, and demographics through multiple personality tests. A new, more comprehensive personality test is set to be released soon. Data analysis will combine demographic, personality, and primal view data to understand aesthetic preferences, with future studies planned to explore how various factors define aesthetics.

Upcoming Features 🚀

Several exciting features are on the horizon, including tools for image captioning, style randomization, higher resolution upscalers, and a character creator tool. Surveys will be expanded to include non-Midjourney users, and new user interaction features like polls and hive mind will be introduced to gather feedback and prioritize feature development.

Future Plans 🌐

To handle the growing demands of feature development and scaling, Midjourney plans to expand its team, focusing on small, self-organized teams for efficient progress. The long-term vision includes real-time, open-world simulators and new hardware interfaces. The team is committed to continuously improving core models for better performance and user experience, balancing simplicity with advanced features to meet diverse user needs.

Miscellaneous 🗂️

Fast hours will expire after 60 days to simplify accounting across multiple countries. A waitlist will be implemented for website access before wider signups. Community engagement remains a priority, with weekly office hours to discuss updates and answer questions, encouraging user participation in feature voting and feedback.

TL;DR;

Personalization Features 🎨: Basic personalization page launching soon; advanced features with multiple profiles planned. Challenges include algorithm complexity and balancing usability.

Model Updates 🤖: Midjourney Version 7 will improve knowledge integration, aesthetics, and prompt understanding. Training delays expected.

Explore Page Improvements 🔍: New algorithm for better image selection, organized feed with "new" and "top" sections, and UI enhancements.

Editor Development 🖌️: Standalone editor with depth references and retexturing. Future features include session-based editing and drag-and-drop functionality.

3D and Video Models 🎥: Scaling up data for 3D models, simple video generation planned, and future VR applications.

Psychometrics and Demographics 🧠: New personality tests and data analysis to understand aesthetic preferences.

Upcoming Features 🚀: Image captioning, style randomization, higher resolution upscalers, and character creator tool. Expanded surveys and user interaction features.

Future Plans 🌐: Team expansion, long-term vision for real-time simulations, and balancing simplicity with advanced features.

Miscellaneous 🗂️: Fast hours expire after 60 days, waitlist for website access, and community engagement through weekly office hours.

Stay tuned for more updates as Midjourney continues to innovate and enhance its platform!

If you’re not subscribed yet to my newsletter Imagine Weekly, I’d be thrilled to welcome you on board!