OpenAI recently launched its latest AI model, GPT-4o. In response, various tech insiders rushed to cover its video demonstration and see its potential.
Now that it’s here, what can you do with it? If you’re skeptical about trying GPT-4o yourself, read further.
READ: GPT-4o is OpenAI’s latest flagship model
This article will discuss some of the newly discovered GPT-4o applications. Note that these may change as the technology continues to improve.
The top 10 GPT-4o applications
- Productivity aid
- Facial expression analysis
- Real-time translation
- Visual assistance
- Mathematics
- Generating video games
- Coding
- Data analysis
- Creating 3D models
- Transcribing ancient texts
1. Productivity aid
The head of OpenAI, Sam Altman, shared his favorite GPT-4o application during an interview in The Logan Bartlett Show on YouTube.
He said GPT-4o helps him maintain his productivity momentum by automating instant responses. Altman said one way he used it is by “putting my phone on the table while I’m really in the zone of working.”
“Then, without having to change windows or change what I’m doing, using it like another channel,” he explained.
“So I’m working on something, I would normally stop what I’m doing, switch to another tab, Google something, click around, or whatever, but while I’m still doing it, to just ask and get an instant response without changing from what I was looking at on my computer that’s been a surprisingly cool thing,” Altman added.
2. Facial expression analysis
GPT-4o’s advanced image recognition capabilities allow it to interpret faces and facial expressions.
Also, it can analyze facial expressions and describe a person’s emotional state in detail. The potential use cases can be astounding, especially for education and customer service.
For example, an educational AI might adjust its teaching style if a student looks bored or disinterested. Moreover, customer service chatbots might be able to read customer faces to see if their sales strategies are working.
3. Real-time translation
GPT-4o can perform real-time translations across multiple languages. The world saw that clearly during OpenAI’s Spring Update 2024.
The livestream translated Italian to English and vice-versa. It facilitated a conversation between Chief Technology Officer Mira Murati and post-training research head Barret Zoph.
The best part is that the model shared side comments and inflected like a real human. As a result, it could serve as a personal assistant that smoothly mediates conversations with foreign language speakers.
4. Visual assistance
The letter “o” in GPT-4o stands for “omni” the Latin word for “all.” It relates to the flagship model’s capability to analyze and provide audio, vision, and text responses in real-time.
However, its most interesting feature is that it can recognize the world around it. Euronews says the model “can describe surroundings, identify objects, and read text, aiding in navigation and interaction with environments.”
READ: Google Astra faces OpenAI’s GPT-4o for AI supremacy
As a result, the news outlet says it could make the world “more accessible to people with low vision by real-time visual assistance.” For example, such a person may ask the bot to identify buildings so that they could reach a specific location.
5. Mathematics
Critics lambasted ChatGPT’s initial launch due to its inability to perform simple mathematical problems. However, mathematics is now one of the GPT-4o applications.
The latest OpenAI livestream shows the model has enhanced reasoning capabilities to answer complex math questions more accurately.
It was so good that it detected a linear equation drawn on paper during the Spring Update. Then, GPT-4o guided OpenAI researcher Barret Zoph in solving it.
Euronews shared an X social media post raving about the AI model’s mathematical capabilities. Soon, it might become an effective math tutor for everyone worldwide.
6. Generating video games
Euronews shared an experiment from Alvaro Cintra, an assistant professor of cybersecurity at Marymount University.
He says he used GPT-4o to generate Python code for a fully functional video game called “Breakout.”
Cintra only provided a screenshot of the game and the request, “Can you please code this in Python?” Surprisingly, the latest AI model performed the task in seconds.
As a result, the cybersecurity expert said GPT-4o has “so many use-cases and faster workflows.”
7. Coding
GPT-4o’s applications go beyond video game development. Figma software engineer Sawyer Hood tested its coding capabilities lately.
He posed his “ultimate LLM test” to OpenAI’s latest large language model: make Facebook Messenger as a single HTML file.
The AI model accomplished the task in 6 seconds. Moreover, Hood shared a silly conversation with the Messenger file to prove the results.
Euronews says there’s another person who tested GPT-4o’s coding mettle. The model produced the website that X user Bryan Marley drew on paper.
8. Data analysis
Data analysis is a highly important part of artificial intelligence development and several industries. It enables people to make sense of the deluge of data from the Internet and other sources.
As a result, people and companies can make sense of this information and use it to serve various purposes.
READ: AI may help us understand what cats say
The CEO and founder of AI media company Superhuman, Zain Kahn, said data analysis used to “take ages in Excel.”
He tested it with GPT-4o and found that it “can generate full-blown charts and statistical analysis from spreadsheets with a single prompt in less than 30 seconds.”
9. Creating 3D models
3D printing is becoming more popular nowadays as 3D printers become more affordable and efficient. It lets everyone with the machine produce any plastic figure imaginable.
However, the technology still has a huge learning curve because it demands an understanding of 3D modeling. Fortunately, it seems GPT-4o might change that.
Euronews shared the GPT-4o application test of X user Min Choi. “I used GPT-4o to create an STL file for a 3D model in 20 seconds on my phone,” he posted.
“Pretty remarkable what you can generate with AI and simple prompts now,” he added. As a result, GPT-4o may facilitate 3D modeling for beginners, convincing more people to use the technology.
10. Transcribing ancient texts
GPT-4o can translate foreign languages, but can it read ancient ones? X user Generative History says it can.
The person submitted a letter with handwriting from the 1800s and asked for a transcription. In response, the AI model returned one with minor errors.
Artificial intelligence has proven that it can decipher ancient languages. Another Inquirer Tech article reported on an AI model that helped translate text from a burnt scroll. Perhaps the researchers who used this model could accelerate their work with GPT-4o.