OpenAI has just announced their latest upgrade to their AI model, known as GPT-4o. This next-generation occurs in real-time and speaks more convincingly like a human, the AI chatbot can interpret and respond to a user’s audio as well as video.
The company has released a series of demos showcasing the advanced capabilities of GPT-4 Omni. In one demonstration, it assists users with interview preparation to ensure they’re aptly presented. In another, it calls a customer service agent to arrange for an iPhone replacement. The chatbot could also participate in light-hearted banter, translating a bilingual conversation in real time, pass judgment on a game of rock-paper-scissors, respond sarcastically when prompted, and even react appropriately when introduced to the user’s puppy.
OpenAI CEO, Sam Altman, expressed his amazement at the bot’s capabilities in a recent blog post: “It feels like AI from the movies; and it’s still a bit surprising to me that it’s real.” With “human-level response times and expressiveness”, this platform is said to be a game-changer.
While a version accepting only text and image inputs was released on May 13, OpenAI plans to roll out the full version in the coming weeks. Intended for both free and paid users on ChatGPT, GPT-4o will be available via their API. The addition of “o” in the name stands for “omni”, denoting a step closer to more natural human-computer interactions.