Under Trump, AI Scientists Are Told to Remove ‘Ideological Bias’ From Powerful Models

The Institute of National Standards and Technology (NISTR), “AI Security”, “AI fairness” and “Ai fairness” and “EU Proofance” and “EU Prosecution”, “EU Proofance and Economy and Economy” and “EU Prosecution and Economy and Economic Competition” (EU Prosecution and Economic Competition) (EU Justice).

The information comes within the framework of an updated cooperative research and development agreement for members of the AI ​​Security Institute consortium sent in early March. Previously, this conclusion encouraged the researchers to contribute to the technical work that could help identify and correct the discrimination model behavior of sex, racial, age or wealth. Such biases can directly affect the end users and damage disproportionate minorities and economically unfavorable groups.

The new agreement marks “to identify and prove the content,”, as well as “to tag the synthetic content”, “to tag the synthetic content” and “tracking synthetic content” and “tracking synthetic content” and “tracking synthetic content” and “tracking synthetic content.” In addition, the American is an emphasis on the first place to put the test tools to develop the test tools “to expand the position of America’s global AI position.”

“Trump leadership eliminated responsibility for security, justice, wrong information and responsibility,” said a researcher with the AI ​​Security Institute, said he did not want to respond.

The researcher believes that it ignores these issues so that it may damage regular users by permitting an algorithms that discriminate on the basis of discrimination based on income or other demographics. “If you are not a technological billionaire, this will lead to a worse future. Wait for the EU’s unfair, discrimination, dangerous and placement,” the researcher claims.

“Wild” says another researcher working with the AI ​​Security Institute in the past. “What does it mean to bloom people?”

Elon muskWhose leaders are currently A controversial effort On behalf of President Trump, he criticized the AI ​​models built by Openai and Google to cut the government’s expenses and bureaucrates. In February last year, a mem wrote in the X, which was labeled “Racist and” Racist “and Openai. Often quotes Even if one of the Google models prevents a nuclear apocalypse, an event, which is argued to be wrong, prevent a nuclear apocalypse and a very difficult scenario. Besides Tesla and EmptyMusk works on Xai, a company that competes directly with Openai and Google. A researcher recently recommended by the Xai has developed a novel technique to change the political pubs of large language models declare string.

The growing body of research shows that there can affect both political biases in AI models liberals and conservatives. For example, Study of the recommended algorithm of Twitter Users published in 2021 showed more prospects based on the right on the platform.

Muskin is the word since January Government Efficiency Department (DOGE) Swept from the US government, creating an environment that thinks that officials are effectively fired, and think that the Trump management is an enemy. As a department of education, some public departments archived documents that celebrate the DEI and deleted. DoGe targeted AISI’s parental organization in recent weeks, Nist. Dozens of workers were fired.