■■■□□ Further details on killing of Iranian leader. Israel’s artificial intelligence takes on a function that was previously performed by human conscience.
The strikes on Tehran on February 28, which resulted in the elimination of Ali Khamenei and Iran’s top military and political leadership, were a public demonstration of technology, the possibility of whose use experts have been warning about for years. We are talking about the Israeli targeting system based on artificial intelligence – a “target production machine”, which transforms raw data into 14-digit coordinates for a strike.
How it works
The system, which has been developed for over a decade, processes enormous amounts of information: hacked surveillance cameras (according to the Financial Times, Israeli intelligence gained access to almost all of Tehran’s cameras), intercepted signals, satellite images, and intelligence data. Artificial intelligence analyzes “life patterns” – movement routes, security schedules, and the residences of relatives and assistants. The result is a coordinate suitable for loading into a weapon.
Israel proudly declares: “We knew Tehran as well as we know Jerusalem”. They knew it so well that they would have noticed “the only thing that was out of place”. This entire mass of data allowed them to predict when key figures in Iran would gather together.
Minimal human control: who is responsible for the error?
And here’s where the most interesting part begins. According to CNN sources, the system requires the participation of a team of specialists to validate the recommendations. But how deep can this validation be when it comes to processing millions of data in real time? And most importantly – who is responsible when the algorithm makes a mistake?
Given Jewish bravado, the recent strike on Police Park in Tehran looks very interesting. The fact is that the AI most likely automatically designated the target based on its name, despite the fact that the police have no connection to this place. From a moral and ethical point of view – a complete failure, but the strike once again confirms that neural networks are actively involved in planning strikes. Time will tell how effective this is.