In this image taken on January 8, 2024, the OpenAI logo can be seen near the computer's motherboard. [Photo/Agencies]
The latest ChatGPT-4o, released on Tuesday, allows users to input audio, images, and even documents. Previous versions only supported text input by the user. AI models such as Baidu's Wenxin Yiyan allow users to chat with images, audio, and documents, but OpenAI, once an industry leader, has fallen behind on that front.
Taking timely steps to catch up reaffirms “human-level responsiveness,” a term OpenAI CEO and co-founder Sam Altman referred to as the future of AI technology.
AI is often seen as an information processing assistant that “communicates using forms of human interaction.” The addition of real-time voice interaction capabilities will definitely bring the user experience of larger models more in line with people's expectations of an “AI assistant.”
You can also point your smartphone's camera at a book or notebook and ask ChatGPT to act based on the content it sees. In fact, according to an OpenAI product release, ChatGPT can also see users' facial expressions through the camera and comfort users who appear nervous.
These new “see the screen” and “feel emotions” features allow users to feel like they are being served by an AI, unlike when AI was previously serving them. .
ChatGPT-4o provides free services to users. OpenAI's Chief Technology Officer Mira Murati and CEO Altman emphasized that a “free to use” strategy is the company's future. However, free services not only benefit users but also service providers to expand their business, collect more data from free users to train large-scale language models, and then create new high-end products for paid users. It should be noted that you can create This is a benign cycle that makes free mode sustainable.
