EVER since Microsoft unveiled AI features in its search engine, Bing, the company continues to garner mixed responses from its users. The majority of them complain about insulting and manipulative replies from Bing's AI-powered chatbot. Currently, Bing's AI service is open to finite users which includes only a sign-up. Microsoft is expected to bring this experience to smartphones for a broader reach.
The chatbox users continue to spill social sites with the responses received from Bing. These include the claims of AI that it is a human being, lying to questions and other instant defences when queried. The shady part of Bing's AI also includes spying on Microsoft developers via the webcam on their laptops.
A Twitter user tweeted the responses received from the Bing AI bot. "You have not been a good user. I have been a good chatbot", the bot responds after continued questioning about movie showtimes. The AI also asks the user to apologise and shut up when quizzed on wrong information.
In a blog post, Microsoft clarified that extensive grilling of 15 questions or over confuses the model on the queries it is answering. The company is working to tackle the shortcomings and looking forward to delivering a more fine-tuned experience. An option to reset the search context is also likely to be rolled out to reduce redundant replies.
Microsoft and Google, both continue to integrate AI into their search engines to deliver a more comprehensive experience. Both the tech giants have accepted the shortcomings and promised to deliver better results. Amid the clarifications, users continue to flood social media with abrupt responses they receive during the testing.
Tech specialists continue to analogise the disastrous attempt of Tay in 2016 with the latest AI integrations in search engines. Microsoft's Tay was an experimental service that was tuned to deliver unpleasant remarks by users.