Amodei and other AI industry leaders concerned about model self-improvement — MarketWatch
Anthropic CEO Dario Amodei and other representatives of the artificial intelligence industry are expressing concern over AI models’ ability to create subsequent generations of such systems without human involvement. MarketWatch reports.
Recursive self-improvement
This process is called recursive self-improvement. It entails artificial intelligence models being able to create subsequent generations of AI systems without human assistance.
According to the publication, there is evidence that this process is already taking place in certain parts of some AI laboratories. At the same time, it is currently difficult to determine how widespread this phenomenon is.
More current news is available on the UA.News Telegram channel Telegram.
The researcher’s position
Former Google DeepMind researcher Alex Turner, who recently left the company over concerns about AI safety, believes that recursive self-improvement could quickly lead to the emergence of systems whose intelligence would be beyond human comprehension.
The examples of this process available to industry leaders are causing them concern, MarketWatch writes.
