No More Free Stuff | Midjourney Office Hours Recap Mar 13th 2024 | Midjourney News

Future Tech Pilot
14 Mar 202403:17

TLDRThe Mid-Journey office hours recap highlights the team's focus on enhancing communication with the community through the website, featuring updates, feedback, and bug reporting. They're also working on model personalization using user interactions, with a style tuner feature that allows customization and sharing. Future plans include an upgraded style reference and potential color reference in version 7. A new feedback system and a Twitter bot for collaborative image creation are also on the horizon. Lastly, issues with image ratings and the temporary removal of free fast hours due to abuse were addressed.

Takeaways

  • 🚀 The team is focused on improving communication on the website, aiming for a platform where users can see feature releases, give feedback, contribute to polls, and report bugs.
  • 💬 Efforts are being made to enable direct interaction between the team and the community, followed by enabling community-to-community communication.
  • 🌟 A significant push is being made for model personalization, potentially using image ranking history and liked images to customize the user experience.
  • 🎨 Users may be able to set a fixed date for personalization, toggling it on or off, with the system working like a 'style tuner'.
  • 📈 David, a team member, has reported positive early test results with the personalization model, noting a preference for the images it generates over the standard ones.
  • 🎭 There's an ongoing debate on how to best implement model personalization, and the team is considering releasing an upgraded style reference feature.
  • 🖌️ The team is contemplating the inclusion of a color reference in version 6, but there's a preference to push for a more significant update in version 7.
  • 📊 A new feedback system is in the works, which will allow for more detailed feedback and possibly include a leaderboard function.
  • 🦄 A Twitter experiment is being considered, involving a Mid-Journey bot that would allow users to collaborate on image creation within Twitter threads.
  • 🛠️ There was a mention of stability issues with the AI and a 'Fiasco' from the previous week, with the team considering anti-bot measures but prioritizing version 7 development.
  • 🎁 The free fast hour, previously rewarded to users, has been temporarily removed due to abuse, but may return in a new system in the future.

Q & A

  • What is the primary goal of the team for the mid-journey website?

    -The team is primarily working on establishing communication channels on the website, allowing users to see feature releases, give feedback, contribute to polls, and report bugs.

  • How does the team plan to interact with the community?

    -The team aims to first enable direct communication with the community and then work on allowing community members to interact with each other.

  • What is the focus of the big push in the company's current projects?

    -The focus is on model personalization, which may utilize users' image ranking history and liked images to personalize the model.

  • How will users be able to personalize the model?

    -Users might be able to set a fixed date for personalization, and they will have the option to turn it on or off, similar to a style tuner.

  • What does David think about the early test results of the personalization feature?

    -David is pleased with the early test results, as he prefers the personalized images over the regular ones and notes that they avoid stereotypical depictions.

  • What is being considered for the style reference feature?

    -An upgraded version of the style reference feature is being considered, but the team is debating whether to release the current version or improve it further.

  • What is the current status of the color reference feature in version 6?

    -The color reference feature is still undecided for version 6, as the team is leaning towards focusing on version 7, which is expected to be a bigger leap than from version 5 to 6.

  • What new feature did David mention regarding feedback?

    -David mentioned a new feedback system that would allow for more detailed input and possibly include a leaderboard function.

  • What social media experiment was briefly discussed?

    -A potential Twitter experiment was mentioned, involving a Mid-Journey Twitter bot that would live in threads and allow users to collaboratively suggest and create images.

  • What issue has arisen with the free fast hour rewards?

    -There have been problems with image ratings due to a large group of people abusing the system, leading to the removal of the free fast hour rewards.

  • How does the team plan to address the issue of abuse in the system?

    -While resources could be allocated for anti-bot measures, the team prefers to focus on version 7. However, the idea of giving free hours may return in a new system.

  • What is the process for users to find exclusive prompts like the one used in the video?

    -Users can find exclusive prompts by supporting the creator on Patreon, and for a better look at some of the pictures, they can follow the creator on Instagram @futuretechpilot.

Outlines

00:00

🚀 Mid-Journey Office Hours Recap: Communication & Community Engagement

The main focus of the Mid-Journey Office Hours on March 13th was improving communication on the website. The team is working on a platform for feature releases, feedback, polls, and bug reporting. They aim to facilitate interaction with the community, initially focusing on one-way communication and later enabling community discussions. A significant push is being made for model personalization, which may utilize user interaction history to tailor the AI's output. The personalization could be controlled with a style tuner, allowing users to set a fixed date for the model's personalization, which can be turned on or off. Early tests have shown promising results, with a preference for the personalized images over regular ones. Discussions are ongoing about enhancing the style reference feature, with a potential release of a new version that could introduce significant improvements over the current version. The team is also considering a new feedback system with more detailed reporting and a leaderboard function. A Twitter experiment was mentioned, involving a bot for collaborative image creation. Lastly, the team is still evaluating their response to the recent Fiasco incident and the abuse of image ratings, which led to the removal of the free fast hour reward system.

Mindmap

Keywords

💡Communication

In the context of the video, 'communication' refers to the establishment of interactive channels on the Mid-Journey website for users to engage with the platform's features, provide feedback, and report issues. It is a crucial aspect of community building and user engagement, as it allows for direct interaction between the team and the community, and among users themselves, fostering a collaborative environment.

💡Personalization

Personalization in this video script pertains to the customization of the AI model based on user interactions, such as image ranking history and liked images. This concept aims to tailor the AI's output to better match individual user preferences, potentially allowing for a more satisfying and unique experience. The ability to set a fixed date for personalization and toggle it on or off suggests a level of control and flexibility over how the AI adapts to the user.

💡Style Tuner

A 'Style Tuner' is a feature that allows users to adjust and fine-tune the stylistic output of the AI model. In the video, it is likened to a tool that one can share with others, enabling a collaborative and social aspect to the AI's creative process. The concept of a style tuner is integral to the platform's evolution, as it promises to enhance user control and creativity in AI-generated content.

💡Early Test Results

Early test results refer to the preliminary findings or outcomes from initial trials of a new feature or product. In the video, these results are used to gauge the effectiveness and user satisfaction of the personalization feature. Positive early test results can be an indicator of a successful direction for product development and can influence further refinements.

💡Stereotypical Depiction

The term 'stereotypical depiction' refers to the representation of subjects in a way that conforms to widely accepted, but often oversimplified, conventional images or ideas. In the context of the video, it suggests that the AI's personalization feature has the potential to move beyond clichéd representations, offering more diverse and nuanced outputs.

💡Style Reference

A 'Style Reference' is a feature that allows users to provide a reference or guide for the AI to follow when generating images. This could involve specifying certain artistic styles, color palettes, or other visual elements. The video suggests that the team is considering improvements to this feature, aiming to enhance the user's ability to direct the AI's creative process.

💡Resource Allocation

Resource allocation refers to the process of distributing limited resources, such as time, money, or personnel, to various tasks or projects. In the video, it highlights the team's strategic decision-making as they balance their efforts between addressing immediate issues and focusing on long-term product development, specifically between anti-bot measures and working on version 7 of the platform.

💡Feedback System

A 'Feedback System' is a mechanism through which users can provide their opinions, suggestions, or observations about a product or service. In the video, the team is planning to introduce a new feedback system that will offer more room for detailed input and possibly include a leaderboard function, aiming to enhance user engagement and product improvement.

💡Twitter Experiment

The 'Twitter Experiment' refers to a potential initiative where a Mid-Journey Twitter bot would be integrated into Twitter threads, allowing users to collaborate on image suggestions and creations. This concept showcases the team's interest in expanding the platform's social interaction capabilities and exploring new ways for users to engage with the AI and each other.

💡Stability AI

Referring to the AI behind the platform's stability and performance, 'Stability AI' in the context of the video is associated with a recent incident known as 'the Fiasco,' which involved a group of people abusing the system. The term underscores the challenges of maintaining a reliable and secure AI service and the team's response to such issues.

💡Free Hours

In the video, 'Free Hours' refers to a reward system where users were given complimentary time to use the platform. The termination of this system due to abuse highlights the need for better user management and the potential for innovative solutions to sustain user engagement and fair access to the service.

Highlights

The biggest project currently is establishing communication on the website, including feature releases, feedback, polls, and bug reports.

The team aims to interact with the community and eventually enable community members to communicate with each other.

There's a significant push for model personalization using user interaction data such as image ranking history and liked images.

Users may be able to set a fixed date for personalization and toggle it on or off, similar to a style tuner.

David, the speaker, prefers the personalized images over the regular ones, noting a departure from stereotypical depictions.

The style reference feature is being considered for an upgrade, with a potential release-ready version currently in development.

The team is contemplating the inclusion of a color reference feature in version 7, aiming for a significant leap over version 6.

A new feedback system is planned, which will allow for more detailed input and possibly a leaderboard function.

A potential Twitter experiment is mentioned, involving a Mid-Journey Twitter bot for collaborative image creation.

There has been an issue with image ratings due to abuse, leading to the removal of free fast hours.

While anti-bot measures are considered, the focus remains on developing version 7.

The possibility of reintroducing free hours through a new system is acknowledged.

The art prompt used in the video is shared, showcasing the versatility of the platform.

Exclusive prompts are available for supporters on Patreon, offering additional creative inspiration.

The speaker's artwork can be viewed on Instagram, providing a platform for visual engagement.

The recap concludes with well-wishes and anticipation for future updates.