Faced with the crazy success of ChatGPT, Google has just unsheathed its competing chatbot called Bard. The conversational robot was presented with great fanfare in Paris. Problem, he made a mistake when answering a question about the James Webb telescope. Bad publicity…
Google takes a dim view of the emergence of ChatGPT. However, the web giant does not intend to be left behind. This is how he has just unveiled his competing chatbot called Bard. This conversational robot will integrate directly into the search engine like the Bing chatbot. It is based on the Lamda language model and promises to provide accessible answers to complex queries.
To boast of its capabilities, Google organized an event in Paris. A presentation conference that allowed us to learn more about how its AI works. While this is not yet available to the general public, we can only refer to the examples given by Google. The opportunity to see that his conversational robot is not yet fully developed…
When the Google chatbot is grossly mistaken
Facing the press, Google presented Immersive View, a function that immerses us in the heart of cities thanks to AI. The giant also explained how artificial intelligence improves Google Translate translations. However, it went downhill when it came to Bard’s strengths. To support its point, Google has unveiled an example of a request that can be submitted to the AI. “What are the latest discoveries made by the James Webb Space Telescope that I can discuss with my 9-year-old?”
Among his various responses, Bard said the telescope had captured the first images of a planet outside the solar system. Erroneous information. The first snapshot of an exoplanet (2M1207b) was taken in 2004 by the Very Large Telescope (VLT). The factual error obviously gave Google bad publicity. The action of its parent company Alphabet also fell 9% on the stock market in the process.
Prior to the error, Vice President Prabhakar Raghavan had pointed out that AI still needed massive testing. “We will combine external feedback with our own internal testing to ensure that Bard’s responses achieve a high level of quality, security, and grounding in real-world information,” said a Google spokesperson, quoted by Reuters.
