[ad_1]
The April 12 news article “With the growth of AI, the Biden administration explores an auditing system” reported that the Biden administration and other government agencies are taking steps to regulate artificial intelligence (AI). This comes in the wake of the quick emergence of AI systems such as ChatGPT. The administration is concerned that ChatGPT and other AI systems could be biased. This is certainly true.
According to the Manhattan Institute’s “Danger in the Machine: The Perils of Political and Demographic Biases Embedded in AI Systems,” ChatGPT has a left-leaning political bias. Through a series of political questions, researchers found that the AI-generated answers were generally left-leaning, including being pro-abortion rights and against the death penalty. It has definitely been proved that ChatGPT is biased, but many still might think it does not require government regulation because the programmers say that the technology is constantly being updated and improved. However, can we be sure that these biases will be corrected? What incentive do programmers have to correct them?
All AI is prone to the biases of its creators. ChatGPT is the most powerful innovation since the internet itself, and it’s important that we regulate a technology with such power.
[ad_2]
Source link
