LogInMckay Wrigley showed how to leverage GPT-4 to create a website with just two lines of text and a simple drawing.
In case you missed it, OpenAI recently presented an improved version of its flagship AI chatbot ChatGPT, featuring new voice and image capabilities that essentially allow the AI to "look" at images, listen to your requests, and offer a voiced response. Thanks to the introduction of GPT-4V (GPT-4 with Vision), the bot can now analyze pictures and communicate with you in a manner that closely resembles human interaction.
While most users are still a few weeks away from experiencing the chatbot's latest features, Mckay Wrigley, a well-known developer of AI tools, who previously managed to incorporate a GPT-4 model in an Apple Watch so that its voice assistant Siri could write code for him, has already had the opportunity to conduct several experiments with the improved ChatGPT, showing how it is possible to leverage its novel ability of sight to create a website out of a rough sketch.
What the author did was draw a simple sketch of a website landing page, upload it to ChatGPT, and ask it to generate code to recreate this image as a Next.js page with Tailwind CSS, emphasizing that the sketch represents a full browser window.
On its first try, it generated the code for a website that closely resembled Mckay's sketch, including smaller buttons, with the only discrepancy being the positioning, which was not entirely precise but still quite close. According to the creator, the experiment was inspired by Greg Brockman's GPT-4 presentation.
Besides that, the developer also managed to turn a picture of a whiteboarding session into code:
And even asked ChatGPT to write code based on a screenshot of a SaaS dashboard:
If you'd like to see more fascinating stuff AI can do, follow Mckay Wrigley on Twitter. In other ChatGPT-related news, a group of 17 authors, including some prominent names such as John Grisham, Jodi Picoult, David Baldacci, and even George R.R. Martin himself, as well as The Authors Guild, have filed a class action lawsuit against ChatGPT developer OpenAI, accusing the company of copyright infringement regarding their fictional works being used to train their AI models.
According to the complaint filed on September 20, the plaintiffs allege that OpenAI downloaded their books from unauthorized e-book sources and utilized them to train GPT 3.5 and GPT 4, the AI models integrated into ChatGPT. You can read the full story here.
Also, don't forget to join our 80 Level Talent platform and our Telegram channel, follow us on Instagram, Twitter, and LinkedIn, where we share breakdowns, the latest news, awesome artworks, and more.