Google has been developing an AI-powered chatbot called Bard to assist users with their everyday tasks. To ensure the quality of responses delivered by the bot, the company has tasked its contract workers to review the responses and rate their accuracy. However, contractors feel they are not given enough time to properly rate the answers and consequently end up simply guessing.
Google enlisted contractors through the firm Appen to improve its chatbot, Bard. Appen employees were not specifically told that they were reviewing responses related to Bard, yet internal documents dated back to February 7, which is around the time when the chatbot was initially unveiled. The contractors in question are involved in the process of evaluating Google’s search algorithms, the relevancy of ads and also flagging websites that are harmful to users. From January, much of their work centred on reviewing the AI prompts, something that seemed to frustrate the contractors.
Although Google asked its full-time employees to review the bot for an approximate time frame of four hours, it concerned contractors that they were not given enough time to rate the correctness of responses and had to guess to get paid. As with all evaluations from an AI, responses should be accurate and coherent, and the raters should know the subject of the conversation – something which a contractor may not have the necessary knowledge on. Google and Appen have not responded to requests for comment.
Google is an American technology company specializing in Internet-related services and products. Through its subsidiaries and services, it provides a wide range of services including cloud computing, digital platform advertising, mobile app development and artificial intelligence. Founded in 1998 by Larry Page and Sergey Brin, Google has grown to become one of the largest and most recognizable technology companies in the world.
Prabhakar Raghavan is the head of Google’s search engineering team. He has been with Google since 2003 and is responsible for improving the company’s search algorithms, driving innovation and scaling the product. During his time at the company, he has driven a number of initiatives, particularly in the areas of natural language processing, artificial intelligence, and machine learning. He holds a PhD from the University of California and is often quoted as an expert in computer science.
