OpenAI CEO Greg Brockman posted from his X account what appears to be the first public image generated using the company’s latest GPT-4o model.
As you may see in the image below, it’s quite convincingly photorealistic and shows a person wearing a black T-shirt with the OpenAI logo writing something in chalk on a blackboard that reads, “Transfer between modalities. Suppose we directly model P (text, pixels, audio) with one large autoregressive transformer. What are the benefits and disadvantages?”
The latest GPT-4o model, which debuted on Monday, improves on the previous family of GPT-4 models (GPT-4, GPT-4 Vision and GPT-4 Turbo) by being faster, cheaper and retaining more information from inputs akin to audio and picture.
It is in a position to do this because OpenAI has taken a different approach than its previous GPT-4 LLM. While these combined many different models together and converted other media akin to audio and images to text and vice versa, the latest GPT-4o was trained on multimedia tokens from the outset, allowing it to directly analyze and interpret image and audio without prior convert it to text.
VB event
Artificial Intelligence Impact Tour: An Artificial Intelligence Audit
Join us when we return to New York on June 5 to have interaction with top executives and delve into strategies for auditing AI models to make sure integrity, optimal performance, and ethical compliance across organizations. Secure your entry to this exclusive, invitation-only event.
Ask for an invitation
Based on the image above, the latest approach is a noticeable improvement over the last OpenAI DALL-E 3 image generation model that debuted in September 2023. I ran a similar prompt on DALL-E 3 in ChatGPT and here is the result.
As you may see, Brockman’s shared image created with GPT-4o significantly improves the quality, photorealism, and accuracy of text generation.
However, GPT-4o’s native image generation capabilities are not yet publicly available. As Brockman alluded to in his X post, saying, “The team is working hard to share them with the world.”