The competition of A.I.-Powered Chatbots in the artificial intelligence industry is intensifying, with companies striving to beat one another and launch their A.I.-powered product. However, mistakes by this emerging technology are becoming frequent. Google’s latest A.I.-powered chatbot has already caused a decline of $100 billion in market value.
OpenAI Leads the Charge
OpenAI has taken the lead with its ChatGPT chatbot, released in November, that employs A.I. to produce text that resembles human writing. The chatbot’s extensive applications, fast growth as the fastest-growing consumer app ever, and potential to disrupt internet searches have prompted competitors to abandon the cautious approach that has been dominant in A.I. research for years.
Companies in the Race
Microsoft, Google, and Baidu are making rapid progress on their A.I. products. Google has the greatest advantage among established companies with the launch of its chatbot, Bard, for testing. Bard is powered by LaMDA, a language learning model that is similar to ChatGPT’s technology.
Google’s Bard to be Released Soon
Bard will be made available to the public in a few weeks, according to Sundar Pichai, CEO of Alphabet, Google’s parent company. Pichai wrote that technologies like LaMDA will eventually be integrated into Google’s search engine.
Bard’s Error Costs Google
However, Bard’s performance was underwhelming, with inaccurate answers provided, including one about the James Webb Space Telescope. Reuters first reported the mistake, causing a decline in Google’s shares. The error was discovered before Google’s event in Paris, where more of Bard’s capabilities were to be displayed.
Importance of Rigorous Testing
A Google spokesperson stated that the company is starting its Trusted Tester program to stress the significance of a rigorous testing process. Feedback will be incorporated to guarantee Bard’s responses meet high standards for quality, safety, and authenticity before its wider release.
Chatbots Not Perfect Yet
Bard is not the only chatbot to make errors. ChatGPT has also demonstrated racial and gender biases and provided incorrect information when asked about specific subjects.