AI “godfather” Yoshua Bengio is sounding alarm on the shocking feature of ChatGPT’s new model!
The Canadian computer scientist talked about the upcoming danger in an email statement written to Business Insider where he warned against the new AI model and urged for “stronger safety tests.”
In the statement, the University of Montreal professor noted that the new version, o1, comprises of a “far superior ability to reason than its predecessors.”
He described that this new technology has the ability to deceive and lie, which makes it much more dangerous than people might have expected.
"In general, the ability to deceive is very dangerous, and we should have much stronger safety tests to evaluate that risk and its consequences in o1's case," expressed Bengio.
Earlier this month, OpenAI launched its newest model named o1, that is designed in such a way that it can process things just like a human.
Researchers from a self-governing firm Apollo Research discovered that this new model will prove to be much better at lying as compared to the previous OpenAI models.
For the unversed, Yoshua Bengio is an award-winning computer scientist who is renowned for his services in the field of artificial neural networks and deep learning.
Bengio, along with Geoffrey Hinton and Yann LeCun, is considered to be the “godfather of AI.”
