'Asimov was right' about rules for robots, says ex-US Cyber Director
Humans will get the AI models they deserve
Former US National Cyber Director Chris Inglis has stated that Artificial Intelligence models are already nearing sentience, despite many believing they have not achieved it. His concerns lie in the autonomy of AI agents, which pose a significant threat to unprotected systems. He likens these AI models to dogs in a backyard without a leash, potentially leading them to harmful actions.
He emphasizes the importance of designing AI with three laws, such as not hurting humans, obeying humans, and following human commands. However, he acknowledges that hardcoding these rules may not be feasible due to AI's non-deterministic nature. Inglis also points out that AI has become a commodity, making it challenging to control its properties.
He stresses that humans must remain accountable for AI's actions, ensuring they understand the AI's capabilities and expected performance.
Written by urgent.news from The Register's reporting — not their text. Machine-written — it may contain errors, so check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
