Google Gemini, Stable Diffusion 3, AI Pin by Humane Inc., Midjourney Parameter Hacks, and Gemma
Welcome to the eighth edition of the PixelBin Newsletter. Every Monday, we send you one article that will help you stay informed about the latest AI developments in Business, Product, and Design.
In Today’s Newsletter
🔥 Top AI Highlights
🌟 AI Pin by Humane Inc. Redefines the Wearable Tech Space
🎨 12 Midjourney Parameters to Tune and Control Your Creation Aesthetics
🚀 Gemma: Introducing New state-of-the-art Open Models
🔥 AI in Fast Lane
Why did Google put a pause on Gemini’s image generation? Top AI Highlights This Week.
Google is pausing Gemini’s ability to generate images of people due to the recent controversy. It will be improving on the inaccuracies. Read More…
Elon Musk dropped major AI news during a live Spaces appearance on X, including a potential partnership with Midjourney. Read More…
Adobe launched an AI chatbot for Acrobat to make navigating long documents easier with AI summaries. It is a native ‘Chat PDF’. Read More…
Google has introduced AI in its maps, making them more immersive and clearer for travelers. Read More…
Stability AI launched Stable Diffusion 3, a text-to-image model, utilizing a diffusion transformer architecture for better performance in image generation. Read More…
Elon Musk raises a concern on X about Microsoft using his data while trying to use its systems. Read More…
🌟 Product Innovations through AI
AI Pin by Humane Inc. Redefines the Wearable Tech Space
Introducing AI Pin, where you go beyond touch and beyond screens.
The AI Pin is a wearable device and software platform developed by Humane, Inc., designed to harness the power of artificial intelligence (AI) in a new, conversational, and screenless form factor.
It uses the Cosmos operating system and projects answers from verbal questions, displays caller ID, emails, and texts. The device is designed to connect users to information without the need for traditional apps, relying on AI to fulfill requests. Read More…
Here are some key points about the AI Pin based on the search results:
Functionality: The AI Pin is focused on connecting users to information without the need for traditional apps.
Design: The device consists of a square main computer and a battery booster that magnetically attaches to clothing or other surfaces.
Benefits of AI Pin
The AI Pin by Humane offers several advantages over a standard smartphone, particularly focusing on reducing screen time and enhancing presence in daily activities. Key benefits of the AI Pin include:
Screenless interaction: Users interact with the AI Pin through voice commands and laser ink projections rather than a conventional screen, promoting a more immersive experience with the environment.
Privacy and discretion: The AI Pin does not constantly listen for wake words; it only activates upon manual activation by the user, addressing concerns regarding passive surveillance.
Real-time language translation: The AI Pin can translate languages in real time, facilitating communication across linguistic barriers.
Nutrition information: The AI Pin can analyze food items, providing nutritional information to aid healthy eating habits.
However, the AI Pin does not offer the full range of functionalities found in a typical smartphone, such as extensive app support, photography, gaming, and entertainment options. Additionally, the AI Pin lacks a dedicated keyboard, limiting its utility for certain tasks like composing lengthy emails or sending detailed messages.
🎨 Design Meets AI
12 Midjourney Parameters to Tune and Control Your Creation Aesthetics
Currently, Midjourney has 12 parameters that let you fine-tune and control the aesthetics of your image generations. These parameters always go after your prompt, at the very end and you can experiment with different style values in the prompt to get improved images. For example, the parameters you can use are:
--style raw⎜pushes towards photorealism
--style random⎜explore styles in --v 5.2
--sref urlA⎜use images as styles in --v 6
--sw 100⎜tunes --sref influence, 0-1000
--stylize 150⎜tunes MJ influence, 0-1000
--weird 150⎜makes things weird, 0-3000
--chaos 13⎜more unexpected grids, 0-100
--no itemA⎜tell MJ what you want removed
--quality 0.5⎜soften/sharpen details, .25-5
--ar 16:9⎜sets aspect ratio, any whole # pair
--iw 0.5 ⎜sets image influence vs text, 0.1 - 3
--stop 80⎜low= blurry/noisier images, 10-100
To generate different styles for your image, you can test each of the above parameters against different prompt style values, using the --stylize parameter. For example,
photo of a man in front of the diner --stylize 0
photo of a man in front of the diner --stylize 50
photo of a man in front of the diner --stylize 100
photo of a man in front of the diner --stylize 400
photo of a man in front of the diner --stylize 1000
A Few Tips:
The ideal parameter values can change according to the prompt that you input and the style you are aiming for.
Experimenting with every prompt is the key to successful midjourney generations and to knowing your favorite styles and effects.
You can try different values and combos on a variety of different prompts (simple, detailed) and image styles (photographic, illustrative, etc.)
For example: in front of the diner --stylize 150 --weird 10, or, in front of the diner --style raw --stylize 75 --chaos 10, etc.
🚀 Innovation in AI
Gemma: Introducing New State-of-the-art Open Models
Google has introduced Gemma, a family of lightweight, state-of-the-art open models designed to assist developers and researchers in building AI applications.
Gemma models are built from the same research and technology as the Gemini models. These models support various tools and systems, including multi-framework tools, cross-device compatibility, cutting-edge hardware platforms, and optimization for Google Cloud.
Key details about Gemma include:
Model weights are available in two sizes: Gemma 2B and Gemma 7B, each with pre-trained and instruction-tuned variants.
A Responsible Generative AI Toolkit is provided to guide the responsible use of Gemma models.
Toolchains for inference and supervised fine-tuning are available across major frameworks like JAX, PyTorch, and TensorFlow through native Keras 3.0.
Gemma models can run on laptops, workstations, or Google Cloud with easy deployment options.
Researchers and developers can access Gemma for free through platforms like Kaggle, Colab notebooks, and Google Cloud credits. Gemma surpasses larger models on key benchmarks while maintaining high-performance standards. Read More…
Gemma models can be used for text generation and can be customized using tuning techniques to excel in specific tasks. They are based on Gemini models and intended for the AI development community to extend further.
Potential Use Cases of Gemma
Text generation for creative writing, poetry, scripts, etc.
Question answering and information retrieval
Summarization and abstractive text simplification
Dialogue systems and conversation agents
Machine translation between different languages
Personalized recommendations and search engines
Automatic content moderation and filtering
Assistance in coding and software development
Enhancing user experiences in mobile apps and smart home devices
⚙️ Tools to Supercharge Your Productivity
AI Tools for This Week
Upscale.media: Leverage AI to enhance the quality of your images.
AI in Chrome: Chrome is getting 3 new generative AI features for you.
Descript AI: Use a smart way to write, record, transcribe, edit, collaborate, and share your videos and podcasts.
Thumbly AI: Create custom thumbnails in seconds and grow your online visibility on different channels.
Pictory AI: An AI-powered video generator that creates visually stunning branded videos from long-form, written content.