Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Disarmingly lifelike: ChatGPT-4o will laugh at your jokes and your dumb hat

May 14, 2024 - arstechnica.com
OpenAI has announced ChatGPT-4o, a chatbot that allows users to interact using real-time audio and video. This new model provides non-verbal cues, making the chatbot feel more human-like. The AI assistant can react to images, mimic vocal intonations, and even alter lyrics in a song. The announcement included several video demos showcasing the chatbot's capabilities, suggesting a significant shift in how we engage with large language models.

ChatGPT-4o's response time has been reduced to 320 milliseconds, a significant improvement from the previous model's two to three seconds. This faster response time allows for more natural conversations, especially in real-time translation scenarios. The advanced vocal capabilities and quicker response time of ChatGPT-4o could lead to a new level of parasocial relationships between the AI assistant and its users.

Key takeaways:

  • OpenAI has announced ChatGPT-4o, a chatbot that allows users to converse using real-time audio and video, which could represent a significant shift in how we interact with large language models.
  • The chatbot's non-verbal cues and human-like responses make it feel more human, which could impact how users perceive and interact with it.
  • ChatGPT-4o's vocal capabilities, demonstrated in several video demos, could lead to a new level of parasocial relationship between the AI assistant and its users.
  • The model's speed of response, reduced from two to three seconds down to 320 milliseconds, could significantly change the way we interact with chatbots, making conversations more natural and less awkward.
View Full Article

Comments (0)

Be the first to comment!