ChatGPT, an interactive chatbot that has taken the tech community and others by storm and created hype on social media platforms over the past few weeks, was touted to replace humans, especially creative writers. However, while there may be demand for it in the long run to assist — not replace — people, it still seems to be early days.
The question-answering artificial (AI) portal provides solutions to most queries posted by users in a conversational tone, giving it a more human touch. ChatGPT has been developed by OpenAI and headed by Sam Altman, former president of Y Combinator, a US-based startup accelerator that funds ideas.
The platform also counts billionaire Elon Musk as a backer. Tech giants like Microsoft, too, have invested in the idea as they look to further their AI offerings.
Also read: Apple’s A17 chip to focus on battery life improvements over power; likely launch with iPhone 15
“ChatGPT is incredibly limited, but good enough at some things to create a misleading impression of greatness. It’s a mistake to be relying on it for anything important right now. It’s a preview of progress; we have lots of work to do on robustness and truthfulness,” Altman had tweeted recently.
“Fun creative inspiration; great! reliance for factual queries; not such a good idea. We will work hard to improve!”
Altman plans to monetise the portal’s offerings, he said on Twitter, but did not provide details. ChatGPT reached the one-million-user mark in about seven days, Altman said, a goal few tech giants can imagine skirting.
OpenAI was targeting a revenue of $1 billion by 2024, up from the $200 million it was projecting in 2023, Reuters had reported citing sources. After ChatGPT wrote essays on several topics, readied business plans for startups and cracked jokes in multiple languages, talks of it replacing Google also made the rounds. But they soon fell flat once users found it was unable to solve basic mathematical operations or give the latest news updates.
“I am a language model trained by OpenAl and do not have access to current events or news. My knowledge is based on a snapshot of the internet as it existed in 2021, and I do not have the ability to browse the internet or access updated information,” ChatGPT’s error message read.
Google, meanwhile, had earlier introduced LaMDA and BERT, both its own technology in the language and conversation space, but recently said there were concerns with conversational AI and hence the company hasn’t opened it up to the public yet.
Also read: Prime Minister’s vision for 2023 is mobile phone exports of Rs 1 lakh crore: Rajeev Chandrasekhar
“We have been concerned … there are still a lot of responsible AI issues with these kind of models. They could sometimes have biases that they have carried, sometimes they could very confidently say wrong things, imagine things instead of sticking to facts. So, which is why Google has been a little more deliberate in not just going ahead and releasing these models to the outside world,” Manish Gupta, director, Google Research India, told some reporters on the sidelines of Google India Day on December 19.
“But yes, there is demand so we are looking at the right balance between courage and boldness and responsibility, or courage and responsibility.”
(With inputs from Jatin Grover)