Google lost $100 billion on the stock market following an error in the texts generated by Bard during his presentation. But it is not the only intelligent chatbot to have made a mistake: the new Bing also made a mistake during the Microsoft conference which lifted the veil on this new function, linked to ChatGPT, of the search engine.
Even though Google Bard is not available to the general public, it has already cost the company $100 billion. And for good reason: in its presentation, Google showed automatically generated answers, which contained false information. While the new Bing has significant differences with ChatGPT, it shares similarities with OpenAI’s chatbot.
However, it is not the only artificial intelligence to be mistaken: the new Bing announced the day before also made mistakes. The first demonstrations of the tool were ” riddled with financial data errors “, writing The Verge. The media relays independent researcher Dmitri Brereton’s analysis which made some discoveries about Bing, but which do not necessarily go in the direction of Microsoft. For the moment, the errors discovered have had no impact on the company’s stock market value.
A new Bing that also makes mistakes
Microsoft presented some possible uses of its new tool integrated into Bing. He can answer questions based on the content of the indexed pages: which makes it possible to create recipes, plan a trip and more generally synthesize search results relating to a query. If it was identified that the AI confused a corded vacuum cleaner with a cordless model, it was also mistaken on financial results. It is precisely on this point that The Vergewinced.
Bing tries to summarize the financial report for the third quarter of 2022 of the ready-to-wear brand Gap: if it indicated a gross margin of 37.4%, the report specifies with an adjusted gross margin of 38 .7%. Bing writes that Gap’s operating margin was 5.9% over this period when it was 4.6%, or even 3.9% after adjustment. For The Vergeit’s very clear : ” the result is a comparison riddled with inaccuracies. »
What Bing’s Mistakes Tell Us About AI
The first thing you learn is that Bing has flaws: it is not completely reliable and the AI behind it is not error-free. For the moment, there is no substitute for human verification, especially in specific and subtle areas such as finance.
The other lesson is that Bing is not based solely on the links it displays in the results. If all the elements come from the pages indexed by the search engine, some are not quoted in the answers. We were able to realize this thanks toa query about the pros and cons of top-selling pet vacuum cleaners“, writingThe Verge. Bing quoted a reference and listed the downsides of the cord, but according to Brereton that’s wrong, it doesn’t have a cable. However, by doing research, we realize that there is indeed a wired version of this portable vacuum cleaner.
The consequence of all this is that Bing uses several sources without listing them completely, which has caused it to confuse two versions of said vacuum cleaner. Another very interesting thing to point out: human verification itself does not necessarily work. Thus, the double check carried out byThe Vergeshows that the researcher was also mistaken.
In short, Microsoft still has some work to do to make the AI-powered version of Bing make fewer mistakes. The firm will be able to rely on the data generated by the first Internet users with access to AI. For the moment, you have to register on a waiting list to hope to be able to use the new Bing. Certain methods of Microsoft relating to this list allowed Bing to be propelled in the ranking of the most downloaded applications of the App Store.
Do you use Google News (News in France)? You can follow your favorite media. Follow Frandroid on Google News (And Numerama).
