The head of Google’s search division has warned that AI chatbots can provide deceptive and misleading responses, according to a recent report. Prabhakar Raghavan, senior Google VP and head of Search, while talking to German newspaper Welt Am Sonntag (via Business Insider) said that AI bots can give convincing but inaccurate answers. He told the German newspaper that this type of artificial intelligence we are discussing can sometimes result in what is referred to as “hallucination.”
“This is then expressed in such a way that a machine provides a convincing but completely fictitious answer.”
Google, which until last year said that it won’t get into AI race citing reputational risk, on Monday rushed to announced its ChatGPT rival-Bard- an experimental conversational AI service. Despite being powered by Google’s own conversational language model, LamDA, the launch of Bard has been criticized by some of Google’s employees as “rushed” and “botched.” This is due to several major slip-ups, including a factual error in the demo video and a roadblock during a presentation.
For the starters, the AI bot messed up on debut when the demo video to elucidate its capabilities had a wrong answer. It had a factual error related to James Webb Space Telescope. The Wednesday event also saw a roadblock when the presenter forgot to bring the phone which was required for the demo video to showcase Bard’s powers.
The launch of Bard which many of Google’s own employees have called “rushed” and “botched” has cost the company nearly $100 billion. The rocky start of Bard sent Alphabet’s shares dipping. The factual error in Bard led to a 9% drop in shares of Google’s parent company making it lose almost $100 billion in market value on Wednesday. This highlights the importance of caution and scepticism when it comes to relying on AI technology for accurate information.
ALSO READ l Top 5 ChatGPT extensions for Google Chrome to boost your productivity
