GPT-4 Has Been Out For 1 Day. These New Projects Show Just How Much More Powerful It Is

On March 14, OpenAI officially announced the launch of the large multimodal model GPT-4. Credit - CFOTO/Future Publishing via Getty Images

The new GPT-4 artificial intelligence software from OpenAI has only been out for one day. But developers are already finding incredible ways to use the updated tool, which now has the ability to analyze images and write code in all major programming languages.

Just hours after its release, several users said they created computer games in less than a minute by simply asking the chatbot to generate code, resulting in near-perfect renditions of Tetris, Connect Four, Snake, and Pong. Other users created a matchmaking service, bedtime stories, a browser extension that translates any webpage into “pirate speak,” and even a tool that can help discover new medications.

OpenAI unveiled the new GPT-4 on Tuesday, saying it can handle “much more nuanced instructions” than the older generation, which captivated users starting in November 2022 with its uncanny ability to generate elegant writing and answer almost any question. But the previous version of Chat GPT relied on an older generation of technology that wasn’t able to reason and learn new things. Its answers were not always correct or appropriate, either.

Read More: The AI Arms Race Is Changing Everything

The new generation, in contrast, relies on a “large multimodal model” that can accept longer text inputs of up to 25,000 words—compared to the previous 3,000 words—and is trained to be safer and more factual, OpenAI said. The public got a sneak preview of the tool when Microsoft last month released the Bing AI chatbot using GPT-4. The tool is so powerful that it can take a photo of a hand-drawn mockup for a simple website and turn it into an actual website using HTML and Javascript code. The new model can even analyze and describe photos, capable of explaining why an illustration of a squirrel taking a photo of a nut was funny (answer: “we don’t expect them to use a camera or act like a human”).

OpenAI hasn’t yet made the image description feature available to the public, but users are already gearing up for its public launch. Joshua Browder, the CEO of the New York-based startup DoNotPay, a chatbot lawyer that helps people with their legal issues, says he plans on harnessing the power of GPT-4’s image feature so users can scan their medical bills and have the chatbot identify errors, compare prices to other hospitals, and automatically draft a legal defense under the No Surprises Act, which protects medical patients from surprise bills.

“The new GPT-4 has almost a 5x improvement on earlier models,” Browder says. “This latest one is sophisticated enough to draft a lawsuit. Earlier models didn’t have that level of sophistication—they would make things up, they would pretend to know things, which we call hallucination. But we’re not really seeing that with GPT-4.”

The buzzy launch of OpenAI’s fourth generation AI software capped months of hype and anticipation over the powers of this technology, which many industry leaders believe has created a fundamental technological shift similar to the creation of web browsers in the early 1990s.

On Tuesday, companies all across the U.S. began coming up with ways to integrate GPT-4 into their products. Duolingo, the language-learning app, announced it would now be able to share more personalized explanations for mistakes powered by AI, which might include “signature, playful and quirky commentary.” At Stripe, the electronic payment platform, GPT-4 is already being used to provide users with accurate answers about customer support and fraud detection. Financial services firm Morgan Stanley is also using GPT-4 to streamline internal technical support processes. Even the government of Iceland is working with OpenAI to help preserve the Icelandic language.

“The speed at which it’s able to program and be creative is pretty staggering,” says Josh Pigford, the founder of Maybe, a financial planning firm that plans to use GPT-4 to quickly extract credit card transaction data for users. “It replaces what an engineer could potentially do.”

In the medical profession, doctors are even considering using GPT-4 for patient consultations. Anil Gehi, an associate professor of medicine and a cardiologist at UNC-Chapel Hill, described to the chatbot the medical history of a few patients he had seen earlier that presented complex medical cases. In just seconds, the chatbot gave him a perfect answer on how he should have treated the patient using all the correct medical terminology. “It very cogently thought through those clinical scenarios and laid out options to consider,” Dr. Gehi says. “It was very impressive.”

“It’s not going to replace a human doctor,” he adds. “But it’s one of those things we can use alongside our work.”

With its wide display of knowledge, the new GPT has also fueled public anxiety over how people will be able to compete for jobs outsourced to artificially trained machines. “Looks like I’m out of job,” one user posted on Twitter in response to a video of someone using GPT-4 to turn a hand-drawn sketch into a functional website. “Idk if i should be excited or scared,” another wrote.

Read More: He Used AI to Publish a Children’s Book in a Weekend. Artists Are Not Happy About It

The nonprofit organization Common Sense released a statement on Tuesday calling for a national dialogue around kids and artificial intelligence, saying that “the consequences for kids, educators and families—both good and bad—have not been thoroughly considered.”

Others expressed concern that GPT-4 still pulls information from a database that lacks real-time or up-to-date information, as it was trained on data up to August 2022. The time-gap could make trusting the accuracy of what’s online more difficult. “The real breakthrough will occur, however, when an AI system…contains up-to-date information—ideally updated in real-time or, failing that, every few hours,” says Oliver Chapman, CEO of supply chain specialists OCI.

To use GPT-4, users have to subscribe to ChatGPT Plus, a $20 monthly subscription, for premium access to the service. GPT-4 currently has a cap of 100 messages every four hours.