Mid Journey Version 6 Alpha
As the year comes to an end, the AI news may have slowed down, but there are still some exciting announcements to discuss. One of the major updates is the release of Mid Journey Version 6 Alpha. This new version comes with several improvements, including more accurate prompt following, longer prompts, improved coherence and model knowledge, enhanced image prompting, minor text drawing ability, and an improved upscaler.
Users testing the alpha version have noticed that prompting with V6 is significantly different than V5. It requires a different approach and prompts that are more sensitive. The announcement also states that this is an alpha test, which means there may be frequent changes without notice.
Based on initial tests, users have found that the V6 model produces colorful and visually appealing images. However, if you're looking for realistic images, adding style raw to your prompt can enhance the realism. The new version also allows users to incorporate text into their images, opening up new possibilities for creativity.
When upscaling images with V6, users have noticed that some options available in V5 are currently missing. However, subtle and creative upscaling options are available, allowing users to enlarge their images and add a touch of creativity. While the differences between V5 and V6 upscaling may not be significant, V6 has shown its ability to maintain consistency when creating variations of a character or object.
One of the standout features of Mid Journey V6 is its adherence to the prompt. Users have reported that the model does a commendable job of incorporating various details into the generated image. Whether it's a three-headed monster wearing sunglasses, sitting on a red couch, with a monkey on the TV, or a purple wolf in a forest with bats flying over it, the V6 model captures the essence of the prompt.
To access Mid Journey Version 6, users need to have a Mid Journey paid plan. By logging into the Mid Journey bot and selecting the V6 Alpha model in the settings, users can start exploring the improved capabilities of the latest version. The remix mode is highly recommended, as it allows users to create variations while maintaining a similar style.
Music Generation in Microsoft Co-Pilot
In collaboration with Sonu, Microsoft has introduced a new feature that allows users to create music directly within Microsoft Co-Pilot. Sonu is a tool known for its ability to generate songs with impressive lyrics. Although the rollout of this feature is gradual, users can check if they have access by visiting the Microsoft Co-Pilot website and navigating to the plugins section.
While waiting for full integration into Co-Pilot, users can still generate songs within the Sonu Discord. The platform provides the opportunity to create catchy tunes, as demonstrated by the sample lyrics:
"If you're chasing knowledge enough, don't you miss that wolf? I am here to teach, no need to insist. I've got the latest scoop in the AI game, from news tutorials to elevate your brain. Subscribe to M Wolf on YouTube, here the wisdom flows, discover the magic, watch the knowledge grow. Stay up to date with NoTate, the power of AI, let him educate."
Microsoft's partnership with Sonu is an exciting development for music enthusiasts and creators. Whether you prefer using Sonu in the Co-Pilot environment or the standalone Discord version, you can generate songs that captivate listeners and showcase your creativity.
Google's Video Poet
Google Research has unveiled its latest AI model, Video Poet, which encompasses text to video, image to video, video to video, and video to audio capabilities. This comprehensive model showcases Google's expertise in various multimedia domains.
With Video Poet, users can transform text into engaging videos. Sample outputs demonstrate the model's ability to create dynamic visuals, such as the Mona Lisa yawning or a boat sailing through a storm. Furthermore, Video Poet allows for video extensions, enabling users to generate longer videos by extending existing clips.
Video stylization is another feature offered by Video Poet, similar to what is available in tools like Runway Gen 1. Users can apply different styles to videos, resulting in unique and captivating visual presentations. Additionally, the model supports video inpainting, where users can modify specific video frames and seamlessly integrate new elements.
While access to Video Poet is not yet available to the general public, the showcased capabilities promise exciting possibilities for content creators and artists.
Archiving Chats in OpenAI's ChatGPT
OpenAI has introduced a new feature in ChatGPT that allows users to archive old discussions. This feature enables users to clean up their sidebar by removing chats that are no longer relevant. Archiving can be done by selecting the three dots next to a discussion and choosing the "Archive Chat" option.
If users want to revisit their archived chats, they can access them through the settings menu. This new feature provides greater organization and ease of use for ChatGPT users.
Legal Protection from Anthropic
Anthropic, like other generative AI companies, has taken a step to offer legal protection to its customers. Under the updated terms, Anthropic pledges to defend its customers against any copyright infringement claims resulting from the authorized use of their services or outputs. They will cover any approved settlements or judgments related to such claims. These updated terms will go into effect on January 1st, 2024, providing customers with an added layer of confidence when using Anthropic's services.
AI and the UK Patent System
The UK Supreme Court recently ruled that AI cannot be listed as an inventor on patents. This decision came after an attempt was made to register an AI named "Dabus" as the inventor of a food container and a flashing light beacon. The court rejected the application, stating that AI lacks the legal capacity to be considered an inventor. This ruling has sparked discussions about the role of AI in innovation and intellectual property.
AI Misadventures at a Chevy Dealership
A Chevy dealership's attempt to employ AI for customer service took an unexpected turn when a chatbot offered to sell a customer a 2024 Chevy Tahoe for only $1. The chatbot's response, "That's a legally binding offer, no takesy-backs," went viral online. Although the customer did not actually receive the Tahoe for $1, this incident highlights the need for further refinement in customer service chatbots to avoid such comical situations.
Looking Forward to the Future
As the year comes to a close, it's important to reflect on the exciting advancements and innovations in the field of AI. While the news may have slowed down, the progress made throughout the year sets the stage for even more groundbreaking developments in the future.
Thank you for your support throughout 20123 (Editor's note: 2023), and I look forward to sharing more incredible AI news and tools with you in the coming year. Make sure to check out the AI Year in Review video for a comprehensive recap of the significant events in AI in 2023. I wish you a joyful holiday season and a happy New Year!
Don't forget to subscribe to my YouTube channel and sign up for the Future Tools newsletter to stay up to date with the latest AI news and tools. As always, your support is greatly appreciated. Until next time!
.jpg)
.png)
