OpenAI, the creator of ChatGPT backed by Microsoft, has officially announced that the chatbot can now access the internet to provide users with real-time information.
Previously, this artificial intelligence-powered system was trained using data only until September 2021. This new capability allows some premium users to ask the chatbot questions about current events and access up-to-date news. OpenAI has plans to extend this feature to all users, including non-paying users, shortly.
ChatGPT goes online
Earlier this week, OpenAI also revealed that ChatGPT will soon be able to engage in voice conversations with users. ChatGPT and similar systems leverage extensive data to generate responses that closely resemble human interactions. These advancements are expected to alter how people seek information online significantly.
Until now, ChatGPT’s “knowledge” has been static, drawn from the internet’s content as of September 2021. It could not browse the web in real-time. For instance, if asked about the most recent earthquake in Turkey or the current status of Donald Trump, it would respond with, “‘I’m sorry, but I cannot provide real-time information.”
The inability to offer information on recent events had deterred some potential users. Tomas Chamorro-Premuzic, a professor of business psychology at University College London, pointed out that this new functionality would make ChatGPT a source of the latest news, gossip, and current events.
He noted that the implication is that it will absorb many questions and inquiries that would otherwise go to search engines or news outlets.
How accurate will an up-to-date ChatGPT be?
However, Chamorro-Premuzic cautioned that using ChatGPT as a search tool has potential downsides. While it offers quick responses to pressing questions, there are concerns about the accuracy and reliability of the information provided, especially if ChatGPT doesn’t cite its sources.
OpenAI has already faced scrutiny from US regulators regarding the risk of ChatGPT generating false information. The Federal Trade Commission (FTC) sent a letter earlier this year asking for details on how OpenAI addresses risks to individuals’ reputations. In response, OpenAI’s CEO committed to working with the FTC.
Why was ChatGPT’s free version so late to the internet?
Several factors contributed to the delay in allowing ChatGPT to search the internet for current information, including computing costs and concerns about accuracy, privacy, and ethics. ChatGPT has been cautious about providing real-time data to avoid the potential introduction of inaccuracies and the risk of accessing copyrighted content without permission.
These new functionalities of ChatGPT underscore the significant dilemma facing the AI sector. AI systems need more freedom and flexibility to be truly useful, but this also increases the risk of misuse and the potential for inaccuracies and misinformation.
