What is AI chatbot phenomenon ChatGPT and could it replace humans?

The tool has impressed experts with its writing ability, proficiency at complex tasks and ease of use

What is the AI bot ChatGPT?

ChatGPT is a prototype dialogue-based AI chatbot capable of understanding natural human language and generating impressively detailed human-like written text.

It is the latest evolution of the GPT – or Generative Pre-Trained Transformer – family of text-generating AIs.

Who made it?

The new AI is the latest chatbot from the Elon Musk-founded independent research body OpenAI foundation.

Musk co-founded the startup with other Silicon Valley investors including technology venture capitalist Sam Altman in late 2015, saying that the research centre would “advance digital intelligence in the way that is most likely to benefit humanity” according to a blog post at the time.

The Twitter CEO has since left the board and distanced himself from the company, tweeting on Sunday that after he “learned” that OpenAI was accessing the platform’s database for “training”, he put a pause on it.

“Need to understand more about governance structure [and] revenue plans going forward,” he said. “OpenAI was started as open-source and non-profit. Neither are still true.”

How does it work?

Trained by AI and machine learning, the system is designed to provide information and answer questions through a conversational interface.

The AI is trained on a huge sample of text taken from the internet.

OpenAI said the new AI was created with a focus on ease of use. “The dialogue format makes it possible for ChatGPT to answer follow-up questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests,” the research body said in a statement last week.

How can it be used?

Early users have described the technology as an alternative to Google because it is capable of providing descriptions, answers and solutions to complex questions including ways to write code, and solve layout problems and optimisation queries.

Real-world applications could include generating content for websites, answering customer inquiries, providing recommendations as well as creating automated chatbots.

Sam Altman, OpenAI CEO, said the system was “an early demo of what’s possible”. “Soon you will be able to have helpful assistants that talk to you, answer questions, and give advice. Later you can have something that goes off and does tasks for you. Eventually you can have something that goes off and discovers new knowledge for you.”

Could ChatGPT replace humans?

There has been speculation that professions dependent upon content production could be rendered obsolete, including everything from playwrights and professors to programmers and journalists.

In the days since it was released, academics have generated responses to exam queries that they say would result in full marks if submitted by an undergraduate, and programmers have used the tool to solve coding challenges in obscure programming languages in a matter of seconds.

The ability to generate human-like written text has prompted suggestions that the technology could replace journalists.

However at its current stage, the chatbot lacks the nuance, critical-thinking skills or ethical decision-making ability that are essential for successful journalism.

Its current knowledge base ends in 2021, rendering some queries and searches useless.

ChatGPT can also give entirely wrong answers and present misinformation as fact, writing “plausible-sounding but incorrect or nonsensical answers”, the company concedes.

OpenAI says that fixing this issue is difficult because there is no source of truth in the data they use to train the model and supervised training can also be misleading “because the ideal answer depends on what the model knows, rather than what the human demonstrator knows”.


Samantha Lock

The GuardianTramp

Related Content

Article image
ChatGPT: what can the extraordinary artificial intelligence chatbot do?
Ask the AI program a question, as millions have in recent weeks, and it will do its best to respond

Ian Sample Science editor

13, Jan, 2023 @4:23 PM

Article image
ChatGPT can tell jokes, even write articles. But only humans can detect its fluent bullshit | Kenan Malik
It has been hailed as the AI program that could spell the end of search engines, but we should beware putting our trust in a machine

Kenan Malik

11, Dec, 2022 @9:00 AM

Article image
‘I want to destroy whatever I want’: Bing’s AI chatbot unsettles US reporter
NYT correspondent’s conversation with Microsoft’s search engine leads to bizarre philosophical conversations that highlight the sense of speaking to a human

Jonathan Yerushalmy

17, Feb, 2023 @9:59 AM

Article image
‘What should the limits be?’ The father of ChatGPT on whether AI will save humanity – or destroy it
Sam Altman is among the most vocal supporters of artificial intelligence, but is also leading calls to regulate it. He outlines his vision of a very uncertain future

Alex Hern

07, Jun, 2023 @5:00 AM

Article image
A lawyer got ChatGPT to do his research, but he isn’t AI’s biggest fool | John Naughton
The emerging technology is causing pratfalls all over – not least tech bosses begging for someone to regulate them

John Naughton

03, Jun, 2023 @3:00 PM

Article image
The existential threat from AI – and from humans misusing it | Letters
Letters: Roger Haines writes that there is no evidence of AI sentience, while Prof Paul Huxley recalls Asimov’s laws of robotics, and Phyl Hyde says fears of AI are being overblown

02, Jun, 2023 @4:16 PM

Article image
What is GPT-4 and how does it differ from ChatGPT?
The most powerful AI model yet from OpenAI, can tell jokes and write bar exams – but can it also cause harm?

Alex Hern

15, Mar, 2023 @5:23 PM

Article image
Colombian judge says he used ChatGPT in ruling
Juan Manuel Padilla asked the AI tool how laws applied in case of autistic boy’s medical funding, while also using precedent to support his decision

Luke Taylor in Bogotá

03, Feb, 2023 @2:53 AM

Article image
The Guardian view on ChatGPT: an eerily good human impersonator | Editorial
Editorial: Artificial intelligence is not artificial consciousness – but it still needs to be regulated to keep people safe


08, Dec, 2022 @6:25 PM

Article image
Italy’s privacy watchdog bans ChatGPT over data breach concerns
Measure is in place ‘until ChatGPT respects privacy’, says Italian Data Protection Authority

Dan Milmo and agencies

01, Apr, 2023 @7:43 AM