We asked ChatGPT to tell us a story about Gritty. It delivered in seconds.
“It was a typical day in Philadelphia, and Gritty was cruising down Broad Street on his roller skates.”
A viral chatbot that uses artificial intelligence to respond conversationally to literally any prompt continues to trend online as users try to understand its uses and limitations.
ChatGPT instantly responds to writing prompts with answers that sound like they came from a real-life person, not a computer scraping the internet (which is basically what it’s actually doing).
The tool, which was created by artificial intelligence company OpenAI and launched earlier this month in prototype mode, is currently free to use. In turn, internet users have been putting it to the test. From letters from Santa and rom-com scripts to workflow hacks, ChatGPT prompts and responses are being shared on social media.
To best explain how the service works, we put it to the test in terms all Philadelphians can relate to: Gritty.
‘Your boy Gritty’
On Friday, I entered the following prompt into ChatGPT’s system: “Write a funny story about Gritty using Philadelphia tropes and slang.”
Instantly, the system began spitting out sentences. ChatGPT wrote its 13-paragraph story in about a minute. Here it is in full:
ChatGPT serves as a dictionary, Google replacement, storyteller, and more
Twitter users have taken to their feeds to show clever applications of ChatGPT. Some users say it could take Google’s place as the top search engine since the chatbot is able to answer prompts instantly and directly. Industry experts say that likely won’t happen any time soon.
Google executives responded this week by saying that Google’s own AI tech is just as capable as OpenAI’s, but the company needs to move at a more “conservative” pace because of the “reputational risks,” CNBC reported.
Here’s a look at some other ChatGPT prompts in practice:
One user asked it to write song lyrics in the style of Taylor Swift.
Another asked it to translate Gen Z slang into something they could understand.
Inc. technology start-up reporter Hillel Fuld suggested using it as leverage at work. Fuld said the chatbot could be used as a smart-editor, suggesting where to trim down copy to fit a strict word count by pasting the text of your post and saying something like “Shorten this text to 2200 characters while maintaining the tone of the post.”
On TikTok, some users have used the chatbot to write Excel formulas for specific tasks.
The chatbot’s popularity is undeniable. Last week, OpenAI CEO Sam Altman said in a tweet that demand was “higher than expected” and they were working to add more capacity.
Concerns over AI technology
AI technology can seem harmless but critics warn that the services can come with ramifications — including devaluing human labor and safety issues.
» READ MORE: Lensa is the app behind your friends’ new profile pics. Here’s what artists have to say about it.
As reported by NBC, there is public concern that bad actors could use the chatbot for harm, asking the service to explain things like how to design a weapon. OpenAI said it has guardrails in place to limit responses to dangerous requests, but said the system isn’t foolproof. Some users have already posted about finding workarounds. Critics also say the chatbot can be easily tricked. One Twitter user said they tricked the chatbot into explaining how to build a nuclear bomb.
ChatGPT’s responses by nature feel authoritative, but it doesn’t mean they’re accurate. OpenAI said that even if a response looks legitimate, it could be nonsensical or incorrect. For that reason, relying on the service at work could be risky.
AI technology continues to trend
Despite its drawbacks, AI technology continues to lure curious users in as it becomes more accessible.
Over the last month, apps like Lensa — an app that uses AI technology to morph photos into stylized images — surged in popularity. Other platforms like DALL-E and My Heritage AI, image generators, have also attracted attention.
Creatives say that while the tools can be fun and tempting to try, it’s important to consider the source of a product. Artists, writers and others in creative fields fear that AI tools will begin replacing human labor for half-baked results.
How can I try ChatGPT for myself?
For now, during its feedback period, ChatGPT remains free. It’s unclear how long the feedback period will last or what OpenAI may be looking to eventually charge for the service.
On Twitter, OpenAI CEO Sam Altman said the company would have to “monetize” ChatGPT “eventually.” He called the costs to run the tech “eye-watering.”
In the meantime, here’s how to try ChatGPT:
Visit ChatGPT’s website: chat.openai.com/chat.
Log in or sign up for the service. You can start an account from scratch or use an existing Google or Microsoft account.
Enter your prompt in the text field. If you’re not happy with the first result, you can ask the chatbot to try again.