Here is your guide to ChatGPT, the trending chatbot that everyone is talking about, if you’re still unclear on what it is.
Even Insider writers have played with the program since OpenAI published its popular chatbot ChatGPT in November, simulating news stories or messaging potential dates.
The informal nature of the chats with the bot can remind older millennials of the experience of conversing online because they were raised with IRC chat rooms, a text-based instant messaging platform. Yet, ChatGPT, the most recent advancement in technology known as “big language model tools,” does not “think” or communicate with sentience like humans do.
While ChatGPT can explain quantum physics and compose poems at will, a full AI takeover isn’t quite feasible, according to experts.
“There’s a saying that an infinite number of monkeys will eventually give you Shakespeare,” said Matthew Sag, a law professor at Emory University who studies copyright implications for training and using large language models like ChatGPT.
“There’s a large number of monkeys here, giving you things that are impressive — but there is intrinsically a difference between the way that humans produce language, and the way that large language models do it,” he said.
Large amounts of data and sophisticated computational methods are used by chatbots like GPT to generate predictions about how to connect words in meaningful ways.
They have access to a huge vocabulary and knowledge, and they also comprehend words in context.
This enables them to convey encyclopedic knowledge while imitating speech patterns.
Some software companies have created their own massive language model tools using programs that process human prompts and provide complex responses, such as Google and Meta.
In a ground-breaking move, OpenAI also developed a user interface that allows the general public to directly experiment with it.
Rob Morris, a co-founder of Koko, hastily clarified on Twitter that although users weren’t conversing to a chat bot directly, AI was being used to “help create” responses.
The controversial DoNotPay service’s creator also claimed that an AI “lawyer” would counsel defendants in actual courtroom traffic cases in real time, though he later withdrew that statement due to concerns about its risks. The chat bot, which is powered by GPT-3, claims to help users resolve customer service disputes.
With generative AI tools, some researchers appear to be using more cautious methods.
The effectiveness of technology in the legal field is studied by Daniel Linna Jr., a professor at Northwestern University who collaborates with the charitable Lawyers’ Committee for Better Housing.
He disclosed to Insider that he is assisting with the testing of a chatbot named “Rentervention,” which is meant to support tenants.
That bot currently uses technology like Google Dialogueflow, another large language model tool. Linna said he’s experimenting with Chat GPT to help “Rentervention” come up with better responses and draft more detailed letters, while gauging its limitations.
“I think there’s so much hype around ChatGPT, and tools like this have potential,” said Linna. “But it can’t do everything — it’s not magic.”
OpenAI has acknowledged as much, explaining on its own website that “ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers.”