The software giant has been getting rave reviews and an incredible response after it launched the competitor of ChatGPT with a bang. Reports proved how millions rushed to the App Stores to download the feature and become a part of the waiting list.
While the launch seemed to be an overall success, one thing is for sure. Critics may have spoken too soon as the company’s public demo carried out last week featured several small errors.
We don’t think that’s a huge deal, considering they’re small in size and also because the majority of data put in front of viewers was correct. Also, let’s not forget how Google’s AI-powered chatbot Bard ended up making a huge error during its public demo. So that might be something Microsoft could be proud of as it hails the new chatbot to be even better than OpenAI’s ChatGPT.
But what exactly was the error in question? Well, it entailed plenty of fake information regarding an earnings report from leading firm Gap. And that was highlighted by an engineer recently that published an entire post regarding it.
During that particular demo, which may be viewed on the web, the firm asked the new Bing search engine to highlight all the key points of the Q3 earnings report from Gap. That is when the program was seen speaking about it and providing a summary of it as well.
For instance, the new chatbot claimed that the company started off with an opening margin comprising nearly 6% but the real report highlights the figure to be 4.6%. Now, the AI-powered search engine also went on to display the figures for sales growth to be in the lower double digits. Meanwhile, the real report outlined how net sales could go down to the likes of mid-single figures YoY during Q4 of 2022.
In one early part of such a demo, the software giant highlights how the new search engine was asked about the pros and cons related to the top-selling variants of pet vacuums. This is when the search engine was quick to highlight a new result and listed down leading benefits as well as disadvantages linked to such products. Even then, critics familiar with the topic claim it made an error in listing down the product’s short length as a disadvantage. The reason is, it’s actually cordless in design and very portable, which is why it’s so popular and user-friendly.
Above all, it seemed to give out details about products that were the most recommended on the web and not necessarily those that were the best-selling. And that again shows how the chatbot can’t be relied upon too much.
Another interesting error is how the chatbot was quizzed about 90s music and while did manage to outline the right stars of that era with their songs, it opted to give out a vague Answer A for all the answers set out in front of it.
Similarly, on Reddit, users are also trolling Microsoft's new venture for providing misinformed answers regarding the number of EU countries.
Read next: Researchers Suggest AI Chatbots Might Make Disinformation Worse
by Dr. Hura Anwar via Digital Information World
No comments:
Post a Comment