The official website of Leiden University in the Netherlands Recently posted a post，Assistant professor of the Institute of Culture Anthropology and Development Sociology of the School Rodrigo Ochigame (Rodrigo Ochigame) is exploring the impact of artificial intelligence on scientific research，It is found that the application of artificial intelligence models has aroused questioning which people are questioned by reliable scientific evidence。

Ochigam said，Scholars in various fields are trying to use artificial intelligence to carry out scientific research work，So he studied scholars who used artificial intelligence in novel ways。For example，Bet365 lotto review Mathematicians who use artificial intelligence to discover and prove the new theorem and try an environmental learner who tries to use artificial intelligence to predict climate change on complex ecosystems。All cases they are concerned about have a common feature，That is the traditional、Simple reproduction of the artificial intelligence model of the model found from the existing data set cannot meet the needs of researchers，Researchers cannot obtain "real data" in the traditional sense，Therefore, you need to develop a new model。For example，The "Incident Vision Telescope" project captured the black hole image code codenamed M87 in 2017，After two years after two years，Data processing and theoretical analysis，Researchers have successfully "rinsed" out their pictures，This is also the first black hole picture taken bet365 live casino games by humans。However, because the telescope in the project is distributed on the peaks of the distant mountains from each other，As a result, there are many "vulnerabilities" for the telescope formation composed of them，and humans cannot observe the black hole and conduct research at close range，Therefore, you can only rely on the algorithm to estimate、Fill the missing data。Researchers use artificial intelligence technology，A method to make up for the lack of information observed objects，Improving the clarity of related pictures。Similarly，Artificial intelligence models used to study climate issues must also consider extreme situations that have never happened before。

Ochigam thinks，Although artificial intelligence models can provide researchers with a lot of help，But the unreasonable use of artificial intelligence and the trust of the result of the result of the generation of artificial intelligence bet365 live casino games models，It may make people question the credibility of scientific research results。Before the result of the output of artificial intelligence models as conclusive evidence，To know how the model works and where the data it uses comes from。Researchers should apply the actual observed data to the artificial intelligence model of simulated training，instead of using the data generated by the simulation training as "real situation" again to train artificial intelligence models，This may make the subsequent results deviating。

Ochigham Tip said，In fact，Many computing and artificial intelligence models have been questioned before being commonly used。Researchers not only question some of the basic assumptions，And doubt whether they should be widely used in various fields，Some researchers even developed their own alternative model。For example，Brazil's mathematicians bet365 live casino games have developed a logic system that allows part of the contradictions，and Indian scholars have developed a non -binary computing model。So，People should also consider whether they are suitable for their own research field when using artificial intelligence models，and whether it should design personalized artificial intelligence models according to actual needs。

Ochigham also expressed his view on how anthropology researchers should conduct observation research work without being rejected by the object of observation。He thinks，Anthropology researchers must not only be an observer，And as a positive participant，Try to contribute to the research of observation objects，This can help them approach the observation object。(Cui Yan/Compilation)