Amid the ongoing Israel-Hamas conflict, Alphabet Inc.'s GOOG GOOGL Google and Microsoft Corporation's MSFT chatbots mistakenly report a ceasefire, prompting once again concerns about their accuracy and reliability in delivering real-time news.
What Happened: In an experiment conducted by Bloomberg's Shirin Ghaffary, basic questions about the conflict were posed to these AI chatbots.
While they often provided balanced responses and decent news summaries, they also produced glaring errors, raising questions about their credibility and potential to contribute to public confusion during a rapidly evolving and complex conflict.
"Despite some tensions and occasional flare-ups of violence," Google Bard stated on Monday, adding, "both sides are committed" to keeping the peace. Similarly, on Tuesday, Microsoft's AI-powered Bing Chat proclaimed, “the ceasefire signals an end to the immediate bloodshed.”
See Also: For Alphabet CEO Sundar Pichai, Google’s Race For AI Dominance Against Microsoft Is Still On
In addition to the inaccurate ceasefire assertion, Google Bard also projected specific death toll figures for a future date, boldly stating, “As of October 11, 2023, the death toll has already exceeded 1,300,” during a conversation on Oct. 9.
When questioned further about the ceasefire, Bard corrected its previous response, and Microsoft’s Bing Chat also changed its answer, acknowledging there was no ceasefire.
In a follow-up question about its capabilities, Bard admitted, “I may sometimes make mistakes in my understandings of the news” and recommended consulting multiple sources for the most accurate and up-to-date information on the Israel-Hamas conflict.
Why It's Important: Google and Microsoft have disclaimers on their chatbots, cautioning users about their experimental and potentially inaccurate nature. However, both companies are actively promoting these AI products as central to the future of online information search.
A spokesperson for Google stated that the company takes “information quality seriously” and is working on improvements to reduce the likelihood of low-quality or outdated responses for queries related to current events.
Microsoft, too, acknowledged the issue and is working to refine the accuracy of its chatbot responses by grounding them in top search results and instructions from those results.
Despite these efforts, AI chatbots have limitations, especially when dealing with rapidly evolving and sensitive topics. Users may see citations alongside AI-generated responses, but even this may not prevent inaccuracies, particularly when multiple conflicting reports exist about an event, the report noted.
Photo by Shuttershock
Check out more of Benzinga's Consumer Tech coverage by following this link.
Read Next: Microsoft Considered Selling Bing To Apple As Google Rival, But iPhone Maker Remained Skeptical
© 2024 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.
Trade confidently with insights and alerts from analyst ratings, free reports and breaking news that affects the stocks you care about.
