OpenAI officially introduced the new model of ChatGPT, GPT-4o, at its Spring Update event today! The most important information announced is that the artificial intelligence model will be released for free for everyone, not just for premium members.
What can GPT-4o do for free?
The innovations within GPT-4o will be available to users in the next few weeks. It also eliminates the need to sign up for ChatGPT.
You will also be able to use data, code and image tools that allow you to analyze images without paying. Questions arise as to whether ChatGPT is worth paying 20 dollars a month in this case.
OpenAI CTO Mira Murati says the biggest benefit for paying users will be the ability to make five times as many requests to GPT-4o per day compared to the free plan.
More than 100 million people use ChatGPT regularly, and GPT-4o will be significantly more efficient than previous GPT-4 versions. Let’s explain.
GPT-4o will be capable of reasoning over voice, text and image
In this new version, you’ll be able to talk to ChatGPT in real time and do almost everything, just like Siri. However, it seems that this model will be far superior to Siri because it can comment not only on your voice but also on the images on your camera.
One of the biggest innovations of the GPT-4o is its live speech feature. This model has the ability to work voice-to-voice. Instead of transcribing speech first, it can listen directly to the audio.
In a demo of this feature, when OpenAI staff approached the voice assistant by breathing rapidly, he could offer advice to improve his breathing techniques. It even gives her a warning like “you’re not a vacuum cleaner, calm down”.
This model is so smart that you don’t have to wait for it to finish its sentence. You can intervene in real time. It can also recognize emotional states.
You can change OpenAI’s voice to your liking
Another demo, as part of the ChatGPT Voice update, demonstrated the OpenAI voice’s ability to be dramatic and emotional as well as natural. The model can produce different voices, including different tones and the ability to sing.
OpenAI can interpret the live stream from your camera in real time
One of the features of the new ChatGPT is local vision capabilities. This basically means the ability to “see” through the camera on your phone. During the demo, the team showed ChatGPT an equation they were currently writing and asked the AI to help them solve the problem. The model can provide not only answers but also step-by-step guidance.
ChatGPT desktop app is coming
Using a very natural sounding voice and running on a Mac, GPT-4o is now able to view and analyze the code being written. It can also explain what it sees and identify any potential problems. During the demo, the AI can analyze a graph and provide real-time feedback and information.
It can also translate in real time
During another demo, the OpenAI team demonstrated ChatGPT Voice’s ability to be used as a live translation tool. It took Mira Murati’s Italian words and translated them into English, and then translated the English answers into Italian.
Artificial intelligence can understand your emotions from the face you show to the camera
ChatGPT can now detect emotion by examining a face through the camera. During the demo, when a smiling face is shown, the AI understands the person’s emotional state and asks “do you want to share the reason for the good energy?”.
There is no clear release date for GPT-4o yet, but it is said to be released in the next few weeks. What do you think? Please don’t forget to share your thoughts with us in the comments section below.
{{user}} {{datetime}}
{{text}}