OpenAI's ChatGPT Can Now Access Real-Time Information From the Internet

In a remarkable development, OpenAI, the parent company behind ChatGPT, has unveiled a significant update that empowers the chatbot to access real-time information from the internet.

This development marks a pivotal moment in the evolution of AI-powered chatbots, as ChatGPT can now provide users with current and up-to-date information from the vast expanse of the World Wide Web. The announcement was made via the company's platform, X, which was formerly known as Twitter.

ChatGPT has come a long way since its inception. Initially, it was trained on data up to September 2021, which meant that it could not offer real-time information. However, OpenAI has consistently pushed the boundaries of AI development, and the latest update represents a significant leap forward.

OpenAI's decision to equip ChatGPT with internet browsing capabilities is a testament to the company's commitment to innovation and meeting the evolving needs of its users. This feature, known as 'Browse with Bing', is currently available to users of the paid versions of ChatGPT, with plans to expand its accessibility to all users in the near future.

OpenAI had previously experimented with a similar feature, allowing users to access real-time information through the Bing search engine within its premium ChatGPT Plus offering. This feature, introduced in May, garnered attention but was temporarily disabled two months later due to concerns that it might enable users to bypass paywalls on certain websites.

The reimagined integration of internet browsing capabilities in ChatGPT appears to have been designed with a more robust framework to address these concerns and deliver a more secure and user-friendly experience.

Interestingly, ChatGPT's new browsing feature draws parallels with Bard, a chatbot developed and launched by Google in March of the same year. This suggests that OpenAI is keen on competing with Google in the AI market, indicating the increasing importance of AI-powered chatbots in the tech industry.

While the introduction of internet browsing capabilities in ChatGPT offers unprecedented access to real-time information, it also raises concerns about user privacy and the potential dissemination of harmful material, misinformation and copyrighted content. As reported by the BBC, users are required to enable their chat history to activate the new browser plugin, essentially consenting to share their personal data with the AI model.

Alex Hanna, director of research at the Distributed AI Research Institute and former research scientist at Ethical AI at Google, emphasised the risks associated with data privacy. There is a legitimate concern that certain private information could be inadvertently exposed through these AI systems, and the extent of data leakage remains a matter of uncertainty due to the proprietary nature of these companies' data sources.

OpenAI acknowledges these concerns and has indicated that the latest feature will allow websites to control how ChatGPT interacts with them, offering some degree of content moderation and safeguarding against potential misuse.

Hanna also highlighted the potential for ChatGPT to generate hallucinations or disseminate misinformation and inaccurate information. This issue underscores the broader challenge faced by AI tools and search engines in perpetuating racial and gender biases, as evidenced by research, including Safiya Noble's book 'Algorithms of Oppression'.

Noble's research illuminates how search engine results for queries about Black women and white women differ, revealing the presence of racism and sexism within the algorithms that power these platforms. The concern over bias in AI systems is not limited to ChatGPT but extends to the broader AI ecosystem.

Earlier this week, OpenAI introduced voice and image features to ChatGPT, a move that elicited mixed reactions from the public. While some users celebrated the advancements as a step toward AI becoming more human-like and versatile, others expressed reservations.

Critics of these developments raised concerns about the potential ramifications for smaller AI startups and software engineers. They fear that the continued evolution of ChatGPT and similar AI models could disrupt the AI job market and potentially displace human workers in certain roles.

Moreover, there have been legal concerns, with recent lawsuits against OpenAI alleging violations of copyright laws and infringement of intellectual property rights. These legal battles further underscore the complexities and challenges that arise as AI technologies advance.

In parallel to OpenAI's endeavours, Meta, the parent company of Facebook, has also made significant strides in the field of AI. Meta recently introduced its first chatbot, Meta AI, which is trained on the Llama 2 language model. This move signals Meta's commitment to integrating AI chatbots into its most popular products, including WhatsApp, Instagram and Messenger.

What sets Meta's AI chatbots apart is their unique backstories, which result in distinct responses to user queries. This approach adds a layer of personalisation and engagement, making interactions with AI chatbots more dynamic and tailored to individual preferences.