GPT-4o is a revolutionary model that reasons across voice, text, and vision. It can analyze real-time video captured by users on their ChatGPT smartphone apps, although this feature is not yet publicly available. The model responds in real-time and can detect and convey different emotions with its voice, similar to rival AI startup Hume.
Prior to GPT-4o, ChatGPT users had limited capabilities with text-only models like GPT-3.5. The introduction of GPT-4o brings a significant upgrade to free ChatGPT users. They will now have access to web browsing, data analysis, chart creation, and even memory storage. The enhanced model can analyze images and documents uploaded by users and supports more than 50 languages.
OpenAI demonstrated various use cases for GPT-4o during the event. One notable example was its ability to function as a real-time translator app, automatically translating speech from one language to another. The model also excels at understanding and discussing shared images and can generate consistent AI art characters.
While GPT-4o will eventually be available to free ChatGPT users, it will first roll out to paying subscribers. OpenAI plans to introduce GPT-4o to ChatGPT Plus and Team users initially, with availability for Enterprise users coming soon. The company also mentioned that it would not open source GPT-4o or any of its newer AI models, which has raised concerns among critics.
The new GPT-4o model will be available in OpenAI’s application programming interface (API) at half the price and twice the speed of GPT-4 Turbo. OpenAI co-founder and CEO Sam Altman confirmed these details in posts on X during the event.
OpenAI’s mindset about building AI has evolved over time. Initially, the company aimed to use AI to create benefits for the world. However, it now sees its role as creating AI and allowing others to use it to create amazing things that benefit everyone. OpenAI plans to charge for certain services to support its goal of providing outstanding AI service to billions of people.
OpenAI also announced the release of the ChatGPT desktop app, starting with macOS and later expanding to Windows. The app will offer a more natural and powerful experience for users. Over 100 million people are already using ChatGPT, and over 1 million custom GPTs have been created in the GPT Store.
Although the event concluded in just 26 minutes, the unveiling of GPT-4o and the ChatGPT desktop app left a lasting impression. With its advanced capabilities and improved user experience, GPT-4o has the potential to revolutionize AI interaction. It remains to be seen how users will embrace this new technology and whether it will surpass previous versions in terms of power, capability, and naturalistic experience.
