For years we’ve been promised that the rise of artificial intelligence will revolutionize everything from our professional lives to our personal relationships – and we’re finally starting to get a real glimpse of what that could look like.
New “mind-blowing” chatbot technology is taking over the internet with users gushing over its apparent ability to write and respond to anything asked of it. San Francisco-based OpenAI released its latest creation, dubbed ChatGPT, to the public just over a week ago, and people can barely contain their amazement.
The software application is designed to mimic human conversation based on user prompts while harnessing the depths of online knowledge and unfathomable computing power to complete written tasks.
Within a week of ChatGPT’s unveiling, more than a million users had tried it, according to company CEO Sam Altman. And the internet talks about it.
What is ChatGPT and how does it work?
OpenAI says its ChatGPT model, trained using a machine learning technique called Reinforcement Learning from Human Feedback (RLHF), can simulate dialogue, answer follow-up questions, admit errors, challenge incorrect premises and reject inappropriate requests.
Initial development involved human AI trainers providing the model with conversations in which they played on both sides – the user and an AI assistant. The version of the bot available for public testing attempts to understand questions posed by users and responds with in-depth answers resembling human-written text in a conversational format.
Or, as Arvind Narayanan, professor of computer science at Princeton put it: “A tool optimized for bulls**tting”.
It’s been through a bunch of iterations already, but it’s still early days. Its touted real-world applications include digital marketing, creating online content, answering customer service questions, or, as some users have found, even helping write and debug code.
Lonely might flock to chatbot: ‘It’s human’
While the most immediate use cases revolve around low-stakes content creation and potential customer service/information provisioning, this could theoretically also play a role in alleviating loneliness.
Australian-born tech James Allworth, who works as chief innovation officer at Cloudflare and co-hosts popular tech podcast Exponent, likened its potential to the movie Hisin which Joaquin Phoenix’s character is romantically involved with an artificially intelligent virtual assistant.
“That’s another thing that it allows, suddenly you have a guaranteed response,” he said during an episode on Saturday.
“It’s that human and the answers are so compelling it’s starting to feel like you’ve passed the Turing test,” he said, referring to the famous test of a machine’s ability to present the equivalent of intelligent behavior.
“You feel like you are talking to a human and there are a lot of people who are alone and would like that,” he postulated.
Its co-host, renowned tech commentator Ben Thompson, said questions about such apps in the future are “certainly very important”, but likened the chatbot to something that can produce answers that have the “quality of a high school essay with the confidence of a 28-year-old man”.
“Everything is going to be different”
Among those caught up in the excitement this week was Aaron Levie, an American entrepreneur and CEO of enterprise cloud company Box.
“ChatGPT is one of those rare moments in tech where you see a glimmer of how everything is going to be different in the future,” he tweeted.
While many marveled at the potential commercial applications, OpenAI CEO Sam Altman was quick to remind people of its limitations.
“ChatGPT is incredibly limited, but good enough at some things to create a deceiving impression of grandeur,” he tweeted on Sunday. “It’s a mistake to rely on it for anything important right now.”
Nevertheless, he described it as “a glimpse of progress”.
Are there any issues with ChatGPT?
As with many AI-based innovations, ChatGPT is not without its worries. OpenAI acknowledged the tool’s tendency to respond with “plausible but incorrect or nonsensical answers,” a problem it considers difficult to solve.
AI technology can also perpetuate societal biases like those around race, gender, and culture. Tech giants including Google and Alphabet Inc’s Amazon have previously acknowledged that some of their projects that experiment with AI are “ethically risky” and have limitations. In several companies, humans have had to step in and undo the ravages of AI.
Despite these concerns, AI research remains attractive. Venture capital investment in AI development and exploitation companies grew last year to almost A$20 billion.
What does Elon Musk have to do with ChatGPT?
OpenAI was founded as a nonprofit in 2015 by Silicon Valley investor Sam Altman and billionaire Elon Musk and has attracted funding from several others, including venture capitalist Peter Thiel. In 2019, the group established a linked for-profit entity to receive outside investment.
Musk, who remains engrossed in his overhaul of social networking company Twitter, quit OpenAI’s board in 2018 but has spoken about the viral phenomenon, calling it “scary good”.
“That’s another thing that it allows, suddenly you have a guaranteed response,” he said during an episode on Saturday.s using it to “train” the tool.s
You can try ChatGPT here.
Do you have a tip for the story? Email: email@example.com.