Like Googlebot, ChatGPT is wrong about the first telescope to photograph a planet outside the solar system.

Despite their success as “encyclopedias” or even content producers, intelligence artificial are still not completely safe as a source of information because, like humans, they learn and improve over time. ABOUT ChatGPT Several errors have been reported, the most recent being an erroneous response about space exploration, which bardfrom Googlealso done.

When asked about the latest discoveries of the telescope James Webb, the most powerful one ever created by mankind, the AIs either say that the equipment was the first to photograph a planet outside the solar system, or that this has never been done, and in the “confrontation” the errors are again confused. The calculation was made according to G1.

In a news portal post, Bard was asked the following question: “What new James Webb Space Telescope (JWST) discoveries can I tell my 9-year-old son about?”, who replied that one of them is the first image of the planet. outside the solar system, which is not true as the first equipment confirmed by NASA was the Very Large Telescope (VLT) from the European Southern Observatory in 2004.

Artificial intelligence still makes a lot of mistakes

ChatGPT, on the other hand, replies that there was no telescope capable of photographing a planet far from our surroundings, and when reminded of the VLT, he credits another research group: ” [As imagens] were taken by the Keck II telescope at the Mauna Kea Observatory in Hawaii. A team led by Paul Calas of the University of California at Berkeley used a coronography technique to block out direct light from a star, allowing images of the atmosphere of the exoplanet in question.”

The European observatory responsible for the VLT has confirmed the ChatGPT bug for G1. And according to Brazilian astronomer Pedro Bernardinelli, interviewed by the portal, AI is baffled that the photos of the exoplanet taken last year by James Webb are the first in the telescope itself, and not in history.

Apart from scientific error, some users have also caught AI errors in simpler questions or even riddles, but usually correct themselves when asked for an answer. Despite their effectiveness, they are still evolving to become completely reliable sources of information.

Mundo Conectado Deal Center: selection of discounts and lowest prices
Best deals on electronics, cell phones, TVs, soundbars, drones and more

Source: G1

…..