Total reading time around 4 minutes.
Welcome to Visually AI!
🔮AI News This Week
OpenAI Spring Update
OpenAI held its Spring Update last week and there were several mind-blowing reveals, including an AI voice assistant capable of speaking with human emotion.
GPT-4o
But the biggest news was the immediate release of the newest flagship model, GPT-4o (“o” stands for “omni”), open to all users - including FREE plans.
GPT-4o is faster, smarter, and can “reason across audio, vision, and text in real time.”
OpenAI says their goal is to make this technology more inclusive and accessible to all users, with new ChatGPT features available to free users, including:
GPT-4 level intelligence
macOS desktop app
Image interaction
Web integration
Data analysis
File uploads
Free users don’t have unlimited daily access to GPT-4o:
“There will be a limit on the number of messages that free users can send with GPT-4o depending on usage and demand. When the limit is reached, ChatGPT will automatically switch to GPT-3.5 so users can continue their conversations.”
ChatGPT desktop app
The new ChatGPT desktop app is also available to free and paid users, but only available for macOS.
A great feature in the desktop app is the ability to take screenshots of browser tabs, or the entire screen with ChatGPT, instead of uploading images.
In this example, I asked ChatGPT to write code for a Contact Me section of my website from a screenshot of the browser tab.
I changed the video speed 2x, but it is fast without it:
The ChatGPT desktop app is still rolling out to all users.
I saw this message letting me know I could download the app while I was using ChatGPT on the web, a few days ago:
Google I/O
At Google I/O 2024, Google announced a variety of updates to their products, including new generative AI tools and features:
Search is being updated with AI Overviews, multi-step reasoning, new planning capabilities, the ability to ask questions with a video, and AI-organized results pages for specific search categories. (link)
Ask Photos, a new experimental feature in Google Photos, uses Gemini models to make it easier to look for specific memories, recall information, and create highlight galleries.
ImageFX now has editing controls that allow users to add, remove, or change specific elements in their images and will also add Imagen 3, a new image generation model. (link)
Veo, Google DeepMind's newest and most capable video generation model, generates high-quality videos up to 60 seconds in length, and will be incorporated into products like YouTube Shorts in the future. (link) (join waitlist for Imagen 3 and Veo here)
MusicFX has a new feature called "DJ Mode" that helps users mix beats by combining genres and instruments. (link)
MusicFX is easy to use, although you’ll have to give it a few seconds to ‘evolve’ music in DJ mode.
I tested it to generate a cinematic score for a viking drama:
You could have your AI service, tool, or event seen by Visually AI’s community of over 9,700 subscribers:
🚀 This Week’s AI Tools
Wegic: AI web designer and developer powered by the latest GPT-4o model, helps you create and modify websites through chatting with an AI assistant. (link)
CSM: Turn text, photos, or sketches into 3D objects with generative AI. (link)
Optimizer AI: AI sound effects generator for creators, game developers, artists, and video makers. (link)
Modyfi: Design, generate, animate, and more on this design platform built for multidisciplinary designers. (link)
Canva GPT: Use Canva in ChatGPT to help you design logos, social media posts, presentations, and more. (link)
Consensus GPT: Research scientific literature, references, and get simplified explanations inside ChatGPT. (link)
Diagrams GPT: Create mindmaps, flowcharts, workflows, and database & architecture visualization for code inside ChatGPT. (link)
🖼️ Image Prompts
PROMPT: minimalist portrait of a woman's face in profile using white lines, with strands of her hair accented by a soft lavender.
PROMPT: vector coastal city, stunning landscape, graphic design, gold indigo white and black, highly detailed
Thanks for reading, and have a creative week!
Thanks once again Heather for your efforts in keeping us all up to speed, and I'm way happy w/my subscription for same. :)
I'd like to share a sad caveat with regard to ChatGPT for desktop--
My old 2017 MacBook Pro is outside the macOS required to run it on my laptop. :(
pssst... if anyone knows a workaround, please share, and yes, I do have the previous 3rd party ChatGPT desktop app.
Looking forward to playing with their GPT store, but the selection doesn't look any better than Poe, and Poe allows you to pull in multiple LLMs into any chat context, and I just don't see much use for the voice stuff. Its really neat, and every advancement in base models is good, but I don't think this is as much of a milestone as people are making it out to be.