The response received satisfied our request
Posted: Wed Jan 22, 2025 10:29 am
We ask YandexGPT to generate responses to different queries
YandexGPT claims that it will be able to list the netherlands telegram database pros and cons of different things, identify their differences, suggest the best option from the possible ones, and answer complex queries based on available sources.
We conducted an experiment: we entered various information queries into the Yandex search bar and analyzed the responses to them.
Request: good qualities of a person in English for a resume.
Quick answer: Yandex has listed the person's positive qualities with a translation in the form of a list. English words are highlighted in bold to differentiate foreign and Russian words. This makes it easier to navigate the already short answer.
Request: comparison of the cost of Porsche Cayenne and Volkswagen Tiguan.
Quick answer: the neural network returned two positions for each car model with the corresponding price range and a short text description.
Result: the neural network answered our question regarding price comparison. But it was confusing that it found the answer on a foreign resource and could not translate it into the language of the search query. We also noticed that the price ranges in the lines of each car model differ slightly from the price range in the text description. This casts doubt on the correctness of the information received.
YandexGPT claims that it will be able to list the netherlands telegram database pros and cons of different things, identify their differences, suggest the best option from the possible ones, and answer complex queries based on available sources.
We conducted an experiment: we entered various information queries into the Yandex search bar and analyzed the responses to them.
Request: good qualities of a person in English for a resume.
Quick answer: Yandex has listed the person's positive qualities with a translation in the form of a list. English words are highlighted in bold to differentiate foreign and Russian words. This makes it easier to navigate the already short answer.
Request: comparison of the cost of Porsche Cayenne and Volkswagen Tiguan.
Quick answer: the neural network returned two positions for each car model with the corresponding price range and a short text description.
Result: the neural network answered our question regarding price comparison. But it was confusing that it found the answer on a foreign resource and could not translate it into the language of the search query. We also noticed that the price ranges in the lines of each car model differ slightly from the price range in the text description. This casts doubt on the correctness of the information received.