- The Renaissance Times
- Posts
- OpenAI Releases "Majorly Improved" GPT-4 Turbo
OpenAI Releases "Majorly Improved" GPT-4 Turbo
The Renaissance Times
April 10, 2024
Dear Artisan,
Every time OpenAI makes an announcement, we all collectively hold our breath for GPT-5, which could be AGI, and are left disappointed.
Today is no different.
OpenAI announced a new model (more like an upgrade to an existing model), called GPT-4 Turbo. Allegedly, it is “majorly improved,” and will be rolling out in ChatGPT in the coming weeks.
Let’s dive in.
Hot Off The Press
Funding for Generative AI Startups Is Shifting to Applications Over Infrastructure
State Street’s generative AI lets investors chat with their data
Poe introduces a price-per-message revenue model for AI bot creators
US bill proposes AI companies list what copyrighted materials they use
Google Unveils RecurrentGemma, Moves Away From Transformer Based Models
The One Big Thing
Maybe I shouldn’t be so harsh on OpenAI.
Not only is this GPT-4 Turbo, it is GPT-4 Turbo with Vision. This means that you can now upload photos and videos as inputs, and GPT-4 will use its Turbo model to have additional context and formulate better outputs based on this vision.
You can see some of the examples of what people are building below:
GPT-4 Turbo with Vision is now generally available in the API. Vision requests can now also use JSON mode and function calling.
Below are some great ways developers are building with vision. Drop yours in a reply 🧵
— OpenAI Developers (@OpenAIDevs)
6:44 PM • Apr 9, 2024
The most interesting part of this to me is OpenAI’s vision of the future.
They changed the game with their text model, have shown hints of audio and video models, and are now adding advanced visual input to their existing production-ready models as well.
This goes to show that the future of AI is very clearly multi-modal. We will have models that can take in various inputs and, initially will be purpose-drive and task-oriented, but over time will learn more generally and be able to accomplish a wide range of tasks.
As some like to call it: AGI.
The Gallery
7 Deadly Sins as AI Art
I used Midjourney v6 to visualize the 7 deadly sins. The results are intriguing.
1. Wrath
— Artificial World (@ARTiV3RSE)
3:02 AM • Apr 10, 2024
Tools
Must have tools for every artisan to add to their toolkit:
QueryPal: A tool to provide instant answers from company documents on Slack and Microsoft Teams
TypeflowAI: A tool to create forms for businesses
Digest: A tool to aggregate and deliver personalized content from various sources into a single newsletter
Vocalist.ai: A tool to transform vocal recordings into singing and rapping performances
Cliplama: A tool to create faceless videos for social media from text descriptions
Deep Tech
The newest and coolest in the research world that you need to know about:
Revising Densification in Gaussian Splatting
Hash3D: Training-free Acceleration for 3D Generation
Eagle and Finch: RWKV with Matrix-Valued States and Dynamic Recurrence
Reconstructing Hand-Held Objects in 3D
MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies
Closing Thought
Saw someone troll Sam Altman as “Sam Almond” and haven’t stopped thinking about it since
Work With Us!
The AI Renaissance is coming and we are building the best community of the people making it happen.
Contact us to sponsor your product or brand and reach the exact audience for your needs across our newsletter and podcast network.