Artificial intelligence safety researcher warns of existential risk
Anthropic researcher Evan Hubinger stated that due to the speed of the technology, the risk of humanity facing extinction in the next decade exceeds ten percent.
Evan Hubinger, a safety researcher at Anthropic, stated that as a result of the rapid development of artificial intelligence, the probability of eliminating all humans within the next ten years is greater than 10 percent.
Risks Regarding Artificial Intelligence Development
Evan Hubinger, a safety researcher working within Anthropic, emphasized that artificial intelligence is progressing very rapidly. Hubinger stated that he believes the probability of destroying all humans within the next 10-year period is above 10 percent.
Current Models and Superintelligence Concerns
Making a statement on the social media platform X, Hubinger stated that the level of risk stemming from current artificial intelligence models is currently at a low level. However, he noted that he harbors deep concerns that the technology could develop and self-improve in a short time, posing an existential threat to humanity.
Alignment Problem and Model Secrecy Allegations
Hubinger stated that his company is striving to do its best, but they do not yet have a clear plan to solve the alignment problem that may arise at the superintelligence stage. On the other hand, a report in the Financial Times claimed that Anthropic hid its latest model from the UK Artificial Intelligence Safety Institute.