- AI Valley
- Posts
- 🐶 Midjourney V6 is almost here!
Together with
howdy, AI enthusiasts!
Here’s what we got for you today:
✨ Midjourney V6 is almost here!
📍 AI determines location from photos
🌐 Tools, Tweets, Resources and Tutorials
Ready, Set, Go: Today’s news should be about a 3-minute read.
UPDATE
According to Midjourney's latest announcement, we can expect to see V6 coming soon with these changes:
Ultra-sharp detail: Expect crystal-clear, hyper-realistic images.
Potential for videos: Your feedback might unlock video capabilities.
Improved Natural Language Processing: It will interpret text inputs better.
Images will be More Accurate.
More creative variations: Expect a range from bold to subtle changes.
Go big with art: We're talking major artworks, not just tiny thumbnails.
And there's more on the horizon:
A funky new Remix Mode.
Turbo Mode: Hold onto your hats, it's 2–3 times faster!
A fresh pricing tier: With faster hours at a lower price and possible generation of more images per command (16 images instead of 4)
Here are some of the v6 photos.
UPDATE
Stanford students created PIGEON, an AI that identifies locations by analyzing Google Street View images with 92% accuracy for countries and can guess a place within 25 km about 40% of the time.
PIGEON, which uses OpenAI's CLIP neural network, excelled in the GeoGuessr game, beating one of the top players. It was trained on 400,000 images from 100,000 GeoGuessr locations, fewer than other models like DALL-E 2.
TOGETHER WITH OCTAML
OctoAI is the fastest, easiest place to run open-source GenAI on super-performant infrastructure. You can now realize never-before-seen LLM performance with the latest availability of Mixtral 8x7B.
With Mixtral on OctoAI:
4x lower price per token than GPT 3.5 with the flexibility of open-source
Experience the accuracy of the first major MoE-powered (Mixture of Experts) OSS LLM
One unified OpenAI compatible API to consume Mixtral, with the capability of serving a cocktail of models for dynamic generative AI
If you want to explore alternatives to closed source and reduce your dependence on GPT, there’s never been a better time.
Check out the Mixtral benchmarks and sign up for OctoAI for free to put it to the test.
TRENDING TOOLS
👨💻 Sketch2App enables users to effortlessly transform hand-drawn sketches into functional code
🧬 Delphi is a cloning tool designed for influencers, creators, experts, and coaches to create an AI version of themselves, allowing for the scaling of expertise and availability.
🤖 SpeakUp AI is a generative AI podcasting tool that turns textual content into amazing podcasts 10x faster using your own voice.
📧 Cold Mail Bot is an automated tool that helps with cold emailing.
💡 Visual Electric allows users to be inspired by its library of stunning imagery and prompts, and remix their ideas through iteration.
🔄 DeepSwapper is a free face-swapping tool that allows users to quickly and easily swap faces in images with high-quality and realistic results.
TRENDING AI
MICROSOFT SUNO
Microsoft's new music creation app demo:
Microsoft just integrated music creation app Suno into its Copilot AI.
Music, lyrics, all instruments and singing by AI.
A version similar will be in open source soon.
100 billion “new” “songs” soon. I’m sorry.
Prompt: “Make a metal song about baseball”.
— Brian Roemmele (@BrianRoemmele)
Dec 20, 2023
HOT OFF THE PRESS
The new AI image generator promises accurate hands and faces down to the last detail while strictly following user prompts.
Imagen 2 enables users to add new elements to an existing image or to expand the image’s original frame.
The AI, named 'life2vec', analyzes various aspects of a person's life to estimate life expectancy. The tool was tested on 6 million Danish people, predicting with over 75% accuracy who would live at least four years beyond 2016.
It's massively multimodal and can take as input: text, image, depth & optical flow, or a masked video and is one of the first models that generates video + audio!
Russian President Vladimir Putin was asked a question by what appeared to be an AI-generated “deepfake” of himself during a Q&A session with the public Thursday.
TWEET OF THE DAY
THAT’S ALL FOR TODAY
Like what you see? Subscribe Now or Partner With Us
If you have anything interesting to share, please reach out to us by sending us a DM on Twitter: @heyBarsee
What did you think of today's newsletter: Your feedback helps to create better content for you. |