Current AGI (1)AI(Note: Generalized Artificial Intelligence) has been heating up unabated. openAI CEO Altman believes that AGI will be realized sooner than expected with existing hardware; and formerGoogle CEO Eric.Schmidt (name)Instead, it says that once AI begins to improve itself, we should consider a moratorium on its further development.
In an interview with the U.S. television network ABC News on the 17th of this month, Schmidt said, "Eventually you may be able to tell computers, 'Learn everything, do everything.' When the system is able to improve itself, we have to think seriously aboutWhether to unplug it. It's going to be extremely tough and it's going to be very challenging to keep the balance."
Schmidt predicts that AI will gradually evolve from a task-specific assistant to aComplex systems capable of independent decision-making. When AI gets to this point, humans should step in and consider shutting down the system, Schmidt noted. Humans will also need toEnsure that the AI cannot counteract the shutdown systemof the effort. "Theoretically, we'd better have someone who can 'control the power,' as the parable goes."
Musk said in October that AI is "most likely to be great," but theThere are also odds that 10% to 20% could go badHe emphasized that the possibility of AI "going bad" is not zero. He emphasized that the likelihood of AI "going bad" is not zero.
Nonetheless, Schmidt is also optimistic about the positive potential of AI. He said, "Importantly, this powerful intelligence means that everyone will have an assistant like the erudite in their pocket at all times. In addition to note-taking and writing assistants, everyone will also haveWise men like Einstein and Leonardo da Vinci advise their own. This will apply to everyone around the globe."