Today in “the tech we wish was around when we were still in school,” a new AI bot just dropped that’s able to write poetry, explain scientific concepts, write code, and much more.
What happened: OpenAI (the lab behind the DALL-E image generator) made its new AI chatbot ChatGPT available to the public as part of its beta test, and it’s getting rave reviews.
- It’s been heralded as a level-up from previous AI chatbots both because of its ease of use and the complexity of its responses (including the ability to get philosophical).
Why it matters: The potential use cases for ChatGTP go far beyond just fooling around and getting it to imitate your fave podcasters. Its potential scope touches everything from forging academic essays, writing journalistic articles, drafting legal documents, and even coding.
- It’s also so good at answering questions that a Microsoft engineer predicted it's less than three years away from handling queries “much better” than search engines.
Yes, but: As of right now, the bot doesn’t have any information past the year 2021 and has a nasty little habit of generating incorrect (or even completely made-up) information.
- And while there are safeguards to prevent inappropriate, illegal, or unsafe content, people have found ways to get around them—like a poem about hotwiring a car.
What’s next: OpenAI plans to allow companies and developers to create products featuring ChatGTP sometime early next year. Then we’ll see what this precocious robot can really do.