[ad_1]
China’s internet regulator, the Cyberspace Administration of China, has begun examining artificial intelligence. Report Version Financial Times.
It is understood that AI models created by ByteDance, Alibaba, Moonshot, 01.AI and other companies will be tested. Neural networks must undergo a series of checks to ensure compliance with “key socialist values.”
The chatbot will reportedly have to answer a number of questions, including those related to ideology and politics.
These checks may seem like “censorship” and interference with the “natural development” of AI, but if we remember that people have become increasingly dependent on neural networks, some frameworks are still needed.
For example, one can’t help but think of a case in which an interaction with a chatbot led to the death of a person. We are talking about a Belgian who committed suicide after a long exchange with a neural network.
Sometimes, the “suggestions” from neural networks can also be life-threatening. For example, “smart speakers” Alex suggestion The little girl’s “challenge”: Don’t fully insert the charger into the socket and then use a coin to close the contacts on the plug.
One can’t help but think of the sometimes rather scary answers that AI gives. Alice’s answer About the cartoon “Masha and the Bear”, also about Legitimate reasons for pedophilia From the chatbot Gemini from Google.
Photo: Anton Kryatov, ONT
follow us telegraph
[ad_2]
Source link
