The chatbot ChatGPT is entering the next round. The company OpenAI has now presented the latest version, GPT-4o. It can now even interact with its users in real time using a human voice.
The introduction of ChatGPT has created a real boom in the field of artificial intelligence. The chatbot from OpenAI has now been given a new version, GPT-4o.
This can “process audio, image and text information in real time”, according to the company’s website. This also means that ChatGPT can now have real conversations with people.
GPT-4o is the new version of ChatGPT
With the new version GPT-4o (“o” stands for “omni”), OpenAI wants to make the interaction between humans and computers “much more natural”. Users can now add any combination of text, audio files, images and videos to the input. GPT-4o can also combine text, audio and images as required for the output.
The new ChatGPT version is particularly fast. A response to an audio input takes just 232 milliseconds. According to OpenAI, this corresponds to the human reaction time and thus enables actual conversations between humans and machines.
Previous versions of ChatGPT required significantly more time. The average latency time for conversations with GPT-3.5 was 2.8 seconds and 5.4 seconds with GPT-4.
In sample videos, OpenAI shows the possible applications for the new version of ChatGPT. Here, for example, a math problem is solved. However, GPT-4o does not simply give the solution, but provides advice on the possible calculation path.
The input works via smartphone camera, which can be used to transmit the task to ChatGPT. Questions about the calculation method are then asked verbally and ChatGPT also answers them.
New version to remain free of charge
The announcement that GPT-4o will also be available to non-paying users may come as a surprise to many. Until now, OpenAI had only made many additional functions available for the paid subscription.
This does not apply to GPT-4o. However, Plus users receive a message limit that is up to five times higher for the new version of ChatGPT.
GPT-4o is also available for developers in the API as a text and image model. According to OpenAI, GPT-4o is twice as fast, half as expensive and has five times higher rate limits than GPT-4 Turbo.
