OpenAI has launched native image generation in ChatGPT, with the 4o model.
“Image generation has largely been a novelty” Sam Altman said in a livestream Wednesday morning ET, on motivation behind introducing the feature.
“I remember seeing some of the first images come out of this model and having a hard time they were really made by AI,” he added in a post on X.
Google had earlier made native image generation available in Google AI Studio for developers across all regions on March 12.
Native image generation support means significantly improved text generation in images, making it more useful to generate images that are geared toward generating social media posts, ads or wedding invitations.
Such models are nascent start to bigger hopes of competing with graphic designing softwares like Canva.
OpenAI has earlier played around with integrating its specialized diffusion-based image models from the Dall-E line in ChatGPT, to little success.
Changing Stance on Censorship
In a shift from long-standing policy, Altman said people would be able to generate “offensive stuff” if they wanted to with the images in ChatGPT.
“This represents a new high-water mark for us in allowing creative freedom,” he said on X.
“People are going to create some really amazing stuff and some stuff that may offend people” but according to the OpenAI CEO checks and balances are in place to ensure that creating such content isn’t the default setting on the model.
Altman said putting “intellectual freedom and control in the hands of users” is the right thing to do, in what is a complete flip of stance from previous years, when OpenAI heavily censored ChatGPT responses, especially image generation with Dall-E series.
What does it take to achieve financial independence and retire early? Fire Fast by Dzambhala helps you understand and plan it out.
Join the vibrant privacy-ensured Dzambhala community on
Want to give feedback on this story? Write to us.