Imagine if Alexa or Siri could churn out a Shakespearean movie review or Siri could write you a college essay.
Last week, OpenAI made ChatGPT available to the public. ChatGPT is an AI-driven chatbot that converses with users in an uncannily believable manner. Users, including academics and some in the IT industry, have been astounded by its capacity to offer extensive, intelligent, and thorough solutions to queries and prompts – even when they are incorrect.
The tool spread like wildfire. Sam Altman, a well-known investor from Silicon Valley and the co-founder of Open AI, announced on Twitter on Monday that ChatGPT had surpassed one million members. Some well-known tech executives, like Box CEO Aaron Levie, were also drawn to it.
ChatGPT launched on wednesday. today it crossed 1 million users!
— Sam Altman (@sama) December 5, 2022
When a new technology alters the way you think about computing, a specific emotion occurs. By Google. Firefox accomplished this. It was completed by AWS. It was the iPhone. With ChatGPT, OpenAI is accomplishing it, according to Levie on Twitter.
There’s a certain feeling that happens when a new technology adjusts your thinking about computing. Google did it. Firefox did it. AWS did it. iPhone did it. OpenAI is doing it with ChatGPT.
— Aaron Levie (@levie) December 4, 2022
But like other AI-powered technologies, it raises potential issues that could include how it might undermine the creative industries, reinforce prejudices, and disseminate false information.
What is ChatGPT?
To generate its responses, ChatGPT uses a sizable language model that has been trained on a vast amount of web data. It comes from the same company as DALL-E, which produces an almost infinite variety of visuals in response to user input. It also represents the GPT-3 text-subsequent generator’s version.
After registering with ChatGPT, users can ask the AI system a variety of inquiries like “Who was the president of the United States in 1955?” or have it break down complex ideas into language that a second grader could understand. Even open-ended queries like, “What is the purpose of life?” will be covered. or “If it’s 40 degrees outside, what should I wear?
“It depends on the kinds of things you want to undertake. Wear a light jacket or sweater, long pants, and closed-toe shoes if you intend to be outside, ChatGPT said. “If you intend to spend the day indoors, dress comfortably in a t-shirt and jeans.”
Yet some users are developing incredibly original ideas.
How people are using it
One user submitted a letter to have a negative entry removed from their credit report, while another asked the chatbot to rework the 1990s hit song “Baby Got Back” in the style of “The Canterbury Tales” (rather than using a credit repair lawyer). Other vivid instances include asking for advice on decorating your home with fairytale themes and asking a question from the AP English exam (it responded with a 5 paragraph essay about Wuthering Heights.)
The style “allows the tool to answer follow-up inquiries, admit its mistakes, challenge faulty premises, and reject improper requests,” according to a blog post published by OpenAI last week.
The page to try ChatGPT was unavailable as of Monday AM due to “exceptionally high demand.” Please be patient as we scale our systems, the message requested. (It now seems to be online again.)
Possible issues
A number of inquiries were effectively answered by ChatGPT, however, some of the answers were obviously incorrect. Because it is “substantially harmful to the site and to users who are asking or looking for correct answers,” Stack Overflow, a Q&A site for programmers and coders, temporarily banned users from sharing information from ChatGPT.
Beyond the problem of disseminating false information, the tool might endanger certain writing careers, be used to explain dubious ideas, and, like any AI tool, propagate prejudices depending on the data set that it was trained on. If you type a question about a CEO, for instance, you might get a response supposing that person is white and male, for example.
On its website, Open AI stated that “although we’ve taken efforts to have the model reject incorrect requests, it will occasionally respond to damaging instructions or exhibit biased behavior.” We use the Moderation API to flag or remove specific categories of dangerous content, although we anticipate that it will initially produce some false positives and negatives. In order to help us in our continued efforts to improve this system, we are glad to gather user input.
ABI Research’s Lian Jye Su, a research director, cautions that the chatbot is still operating “without a contextual grasp of the language.”
It is relatively simple for the model to provide responses that are true or illogical despite appearing plausible, he claimed. It made assumptions when it should have clarified and occasionally followed harmful instructions or acted in a prejudiced manner. Additionally, it lacks knowledge of particular regions and nations.
However, it also offers a glimpse into how businesses could be able to profit from creating solutions for patient and customer care, as well as more capable virtual help.
Even though the DALL-E tool is free, there is a cap on how many prompts a user can complete before having to pay. The average cost of a ChatGPT session was recently questioned by Elon Musk, a co-founder of OpenAI, on Twitter. Altman responded: “We will have to commercialize it somehow at some time; the compute expenses are eye-watering.”
