DeepMind, the artificial intelligence research unit of American technology giant Google, has developed a new artificial intelligence. “SAFEThis artificial intelligence, called “, is quite remarkable in its properties. Because this artificial intelligence is developed to control the world’s most popular artificial intelligence instruments. SAFE is “Artificial intelligence will take over our jobsIt looks like this will reignite the debate.

SAFE is the foundation of productive artificial intelligence tools such as ChatGPT. to check the results of large language models developed. The system works as follows; artificial intelligence produces a text based on the user’s input. SAFE to check that the information in this text is correct. Google searchIt benefits from. If the results are correct there is no problem, but for incorrect results there is artificial intelligence such as ChatGPT stimulated again.

Research shows that SAFE is better than people

The Google DeepMind team has done extensive research to see how successful SAFE is. This research shows that artificial intelligence 72 percent It showed that it exhibited the same performance as humans. Researchers exploring the issues in which humans and artificial intelligence differ, this time based on the SAFE assessment. 76 percent of It turned out to be more accurate. In other words, the artificial intelligence that controls artificial intelligence has surpassed humans.

SAFE has another advantage over humans. The DeepMind team believes that the cost of owning this artificial intelligence is lower than hiring an employee. 20 times cheaper He discovered that this was so. In other words, a company that gets help from SAFE instead of hiring employees will achieve more accurate results more cheaply. Let’s see when SAFE starts to become widespread. What kind of discussions? will reveal…