The Pentagon is experimenting with advanced AI systems that want to look “a few days ago” to predict the next move from the enemy.
New project dubbed Global Information Superiority Experiment (GIDE)Uses a combination of machine learning algorithms, cloud computing, and submarine sensors to screen large amounts of raw real-time data. The goal is to find even the slightest changes in sensitive areas that may suggest potential problems.
For example, if satellite data suggests that an enemy submarine is about to leave, the system warns a human agent, who can scrutinize its location.
Similarly, if satellite imagery shows that hostile troops are moving, the system predicts where they are likely to be deployed and blocks U.S. forces from hostile activity. Allows you to deploy troops in advance.
The US Northern Command recently completed a third series of futuristic experiments, GIDE 3. In this experiment, all 11 U.S. headquarters and the Pentagon will use a variety of military and civilian sensors to focus on scenarios in which “logistics conflicts” (such as communications on the Panama Canal) can occur. I guessed it. ) Security issues may occur.
At a media briefing last week, General Glen VanHerck, commander of the United States Northern Command (NORTHCOM), said GIDE’s main objectives were “achieving information superiority” and “decision-making superiority.” ..
“I’m talking for days, not minutes or hours,” said General Glenn Vanhelk, who believes in this prediction. AI system It may help the Pentagon take proactive steps, such as strengthening defenses and deploying troops, while at the same time providing an “opportunity” for civilian governments.
“If you can see it days ago, it creates a decision-making space. As a potentially stance commander of operations, the decision-making space is a deterrent option to provide it to secretaries and presidents. To use messaging, you need to use the information space to create a deterrent option. You need to send a message and take further steps as needed to prepare yourself for defeat. “
He said the focus of the latest test was the “two peer rivals” of the United States, without direct reference to China and Russia.
The move is more than four months after the National Security Commission on Artificial Intelligence (NSCAI) warned in a report. The United States was not ready for the AI threat, And soon could be replaced by China as the world’s AI superpower.
“Americans have not yet addressed how serious the artificial intelligence revolution will be on our economy, national security and welfare,” the report said.
“We still need to learn a lot about the power and limitations of AI technology, but we still need to make big decisions to accelerate AI innovation to benefit the United States and prevent AI abuse. “
Created in 2019, NSCAI requires the U.S. President and Congress to make recommendations to advance the development of AI, machine learning, and related technologies to meet national needs in terms of defense and national security. increase. It is led by former Google CEO Eric Schmidt and vice-chaired by former Deputy Secretary of Defense Robert Work.
https://www.computing.co.uk/news/4035332/pentagon-tests-predictive-ai-threats-advance The Pentagon tests predictive AI to proactively identify threats