New AI-powered chatbot made a mistake over the James Webb Space Telescope in an advertisement
Google’s parent company Alphabet lost $100 billion in market value on Wednesday after its new artificial intelligence-powered chatbot,Bard, made an error in a promotional video.
Alphabet shares fell 9 per cent during trading and closed down 7.68 per cent.
Google was showing off its new technology at its Live from Paris virtual event when Bard was asked what a parent could tell their nine-year-old child about the discoveries made by the James Webb Space Telescope.
It responded that the telescope was the first to take pictures of a planet outside Earth’s solar system. But that was achieved by the European Very Large Telescope.
Google launched Bard on Monday to compete with rival ChatGPT, an AI service created by OpenAI.
The service aims to create innovative ways to engage with information, from language and images to videos and audio.
Sundar Pichai, chief executive of Google and Alphabet, made reference to the James Webb Telescope question in a blog post on Monday.
“Bard seeks to combine the breadth of the world’s knowledge with the power, intelligence and creativity of our large language models,” he said.
“It draws on information from the web to provide fresh, high-quality responses.
“Bard can be an outlet for creativity, and a launchpad for curiosity, helping you to explain new discoveries from Nasa’s James Webb Space Telescope to a nine-year-old, or learn more about the best strikers in football right now, and then get drills to build your skills.”
The incorrect answer remains in a gif on the blog post.
The California-based company said it would allow “trusted testers” access to the new service in the beginning. It plans to make it widely available in the coming weeks.
ChatGPT launched in November and became an instant hit, although it led to concerns among some teachers that the program could lead to an increase in plagiarism. It was banned in New York City schools.
OpenAI has since introduced a new tool that can detect whether a piece of writing has been created by AI or a human being.
But OpenAI said it was not totally successful in identifying AI text.
In an evaluation of English texts, the classifier correctly identified 26 per cent of AI-written text as “likely to be AI-written”, but incorrectly labelled human-written text as AI-written 9 per cent of the time.
Meanwhile, Microsoft entered the AI-powered search field this week.
The Bing search engine and Edge web browser will deliver more complete answers to queries, a new chat-style search experience and the ability to generate content, Microsoft said on Wednesday.
Bing runs on a next-generation large language model built by OpenAI that Microsoft described as more powerful than ChatGPT.
Microsoft’s upgrade of Bing, which is live in limited preview on desktop computers and will be available for mobile devices in the coming weeks, takes key learnings and advancements from ChatGPT and GPT-3.5 and is “even faster, more accurate and more capable”, the company said.