Supported by Microsoft, OpenAI's ChatGPT now has internet browsing capabilities to offer users the latest information. The AI system was initially trained using data only until September 2021.
This change allows select premium users to inquire about contemporary events and access news with the chatbot. OpenAI announced that all users would soon access the feature. Earlier this week, OpenAI disclosed that the chatbot will soon support voice interactions with users.
ChatGPT and comparable systems utilize vast data volumes to craft authentic, human-esque replies to user questions. They're anticipated to revolutionize online information search methods.
Can AI will be able to write jokes on current news?
Until recently, the popular chatbot's information was static, sourced from the internet's content as of September 2021. It couldn't access the web in real-time. For instance, if you query the free version about the latest earthquake in Turkey or Donald Trump's current status, it responds, "I'm sorry, but I cannot offer real-time data."
Some prospective users have been deterred by ChatGPT's limitation in addressing recent events. Tomas Chamorro-Premuzic, a professor of business psychology at University College London, remarks, "Without this feature, you'd have to consult Google, Twitter, or your favorite news source. Now, this can serve as a conduit for the most recent news, chatter, and happenings."
"He added, "Thus, a significant consequence is that many questions and queries previously directed to search engines or news outlets will now come here." However, Mr. Chamorro-Premuzic cautioned that utilizing the platform for searches might have both pros and cons.
"It's beneficial for obtaining swift answers to urgent questions," he mentioned, but emphasized the potential for misleading information from ChatGPT without proper sourcing.
"If it doesn't credibly indicate its sources and merely combines various online content, then there are concerns about accuracy. People might mistakenly assume the provided information is trustworthy when it might not be," he explained.
OpenAI has already faced scrutiny from US regulators due to concerns about ChatGPT producing inaccurate information. Earlier in the year, the Federal Trade Commission (FTC) reached out to the Microsoft-supported company, seeking details on how it manages risks associated with individuals' reputations.
In reply, OpenAI's CEO stated that the firm would collaborate with the FTC. Until now, several factors prevented ChatGPT from browsing the internet, one being the computational cost. It's commonly mentioned that each query to OpenAI incurs a few cents in expense.
However, more importantly, the restricted data served as an essential safety measure. This meant ChatGPT couldn't inadvertently reproduce harmful or unlawful content freshly posted online when answering a question.
Since it lacked access, it couldn't propagate misinformation introduced by malicious entities regarding politics or health choices. When queried about the delay in enabling users to access current information, the chatbot itself offered three explanations.
The chatbot explained that creating language models is time-consuming and requires significant resources. Additionally, incorporating real-time data could lead to inaccuracies. Furthermore, there were privacy and ethical issues tied to accessing current information, especially accessing copyrighted content without authorization.
The latest features of ChatGPT underscore a major conundrum in the AI industry. For AI to be genuinely effective, its restrictions must be removed or at least relaxed. However, this very act amplifies the technology's potential risks and susceptibility to exploitation.
Stay tuned and follow us for the latest industry updates and insights.