Former Google CEO Eric Schmidt warned that AI systems may need a “kill switch” if they become too powerful, posing a threat to humanity. As AI advances and becomes increasingly autonomous, it could present new and grave dangers. According to Schmidt, humans will need to step in and consider turning off the system when it reaches a point where it can self-improve.
Schmidt predicts that AI will advance from task-specific agents like Microsoft Copilot to more complex systems capable of making decisions on their own. When this happens, humans will need to ensure that the AI itself cannot counteract efforts to shut it down.
While some experts, including Nobel laureate Geoffrey Hinton and Elon Musk, share Schmidt’s concerns about the potential risks of AI, others believe in its benefits. However, Schmidt emphasizes the importance of regulation to prevent significant damages.
Schmidt notes that governments have a role in regulating AI, and he believes that technologists should not be the only ones making decisions. The U.S. has yet to implement federal-level AI regulation, but some states like California are taking steps to protect industries and regulate deepfakes.
Ultimately, Schmidt’s warning highlights the need for human oversight to ensure that AI is developed and used responsibly, preventing a potential catastrophe.
Source: https://fortune.com/2024/12/16/ex-google-ceo-eric-schmidt-warns-ai-self-improve-unplug-it