OpenAI unveiled GPT-4, its latest AI model. Here are five creative ways people are using it

OpenAI, the company that invented the much-hyped ChatGPT chatbot, has released the latest and “most advanced” version of its artificial intelligence (AI) model.

The GPT-4 language model can “solve difficult problems with greater accuracy”, is more reliable and can handle “more nuanced instructions” than its predecessor system, OpenAI said on its website.

Launched in November 2022, ChatGPT quickly sparked both excitement about its capabilities and criticism of AI systems, including concerns about the ethical use of the program in schools and offices.

Here’s a look at the capabilities of the latest GPT model, its limitations, and five practical ways to leverage AI technology.

What is GPT-4 and what’s new?

According to OpenAI, GPT-4 demonstrates “human-level performance”.

“It’s not perfect, but neither are you. And together, it’s this amplifying tool that makes it easy for you to reach new heights,” OpenAI President Greg Brockman said during an online demonstration Tuesday.

How does it work? Essentially, GPT, which stands for Generative Pretrained Transformer, is an AI model trained to process prompts in order to carry out the tasks at hand, e.g. B. answering questions or generating code.

The biggest update is that GPT-4 can see, understand and react to images. The earlier version was limited to text prompts. The company says GPT-4 can handle up to 25,000 words – nearly eight times more than its predecessor.

The new system is also “controllable,” meaning users can customize it, for example by asking for answers in Shakespearean pirate-style, rather than receiving an answer in ChatGPT’s “classic” firm tone and verbosity. GPT-4 can also calculate tax deductions, compose songs and write screenplays.

However, OpenAI has acknowledged that GPT-4 still has limitations and warned users to be careful. The system is “still not fully reliable” because it “hallucinates” facts and makes mistakes in reasoning, the company said.

Here are five things GPT-4 and its new imaging capabilities can do

It can turn a hand-drawn sketch into a functional website

During the OpenAI live demonstration, Mr. Brockman sketched the design of a simple website on a napkin, photographed his drawing with his phone and uploaded it to the GPT system. Moments later, the technology generated HTML code and turned the sketch into a working website.

It can pass standardized tests

According to a report published on OpenAI’s website, GPT-4 can pass standardized tests such as the SATs and the Bar Exam. For example, GPT-4 received a score in the top 10th percentile during a mock bar exam. In contrast, the previous version was in the lower 10th percentile.

It can help you decide what to cook for dinner based on a picture

The OpenAI website shows an example where you can upload a photo of food in your pantry or fridge to help determine what to eat for breakfast, lunch, or dinner. You can enter the photo with the prompt “What can I make with these ingredients?” and GPT-4 will return a list of recipe ideas.

It can possibly help you find love

Keeper Match, an AI-powered dating site, experimented with using GPT-4 for matchmaking. On Twitter, Chief Executive Officer Jake Kozloski posted a video showing a user uploading two customers’ profile details to see if a match would work.

It can recreate your classic childhood games

If you miss playing nostalgic online games, GPT-4 can program one for you. A Twitter user said they recreated Pong, one of the first computer games of all time, in less than 60 seconds. Likewise, the Snake game was recreated in less than 20 minutes. Another example on TikTok features GPT-4, which recreates the classic arcade game Space Invaders from the 1970s in three minutes.

With reports from The Associated Press.