Wuhan University rule-breaking with AI-controlled satellite experiments: experts

Researchers at a Chinese university last month allegedly handed over control of a satellite to an artificial intelligence (AI) program for 24 hours, showing how far the country will go to find ways to get ahead using AI technology, experts warn.

“Many Americans understandably want to hit the pause button on AI development to sort out the risk issues. China, unfortunately, is roaring ahead, as its 24-hour satellite experiment shows,” Gordon Chang, a China expert, told Fox News Digital.

Researchers at Wuhan University allegedly handed over control of the Qimingxing 1, a small Earth observation satellite, to a ground-based AI program. The program had freedom, with no human orders, assignment or intervention, the South Morning China Post reported. The researchers developed the AI using data from around the globe, creating it not to chat but to take initiative based on its training and growing understanding of natural and human activities.

Lead researcher Wang Mi said that the experiment broke the rules of mission planning, which requires satellites to have specific orders or assignments before taking action.


Space lecture Wuhan

Huo Liang, founder of Jiangsu Shenlan Space Co. Ltd., gives a lecture to students at Wuhan University on China’s Youth Day on May 4, 2023, in Wuhan, China. (VCG/VCG via Getty Images)

During the alleged experiment, Wang’s team observed the satellite as it picked out locations on Earth to make closer observations. The satellite identified an ancient city by the Ganges River in northeast India and home to the Bihar Regiment, which engaged with Chinese forces in the disputed Galwan Valley in 2020, and it also focused on the Japanese port city of Osaka, which occasionally hosts U.S. Navy vessels, according to the SMCP.

A State Department spokesperson told Fox News Digital that the department was aware of reports about the Wuhan University experiment but referred to the university and China’s government for any further details.

“The Communist Party’s only regulation of the technology is to make sure that nobody uses AI to criticize, mock or otherwise undermine its rule,” Chang said. “Because we don’t want to live in a world where Chinese communists dominate AI, we have no choice but to continue development as fast as we can. China can, single-handedly, prevent humanity from adopting safeguards.”

“The bottom line: Chinese communists will do anything, which means we must match them step for step in AI,” he added. “This is not an ideal outcome, but ideal outcomes are not possible.”


Wuhan university experiments

Graduates wave Chinese national flags during the graduation ceremony at Wuhan University on June 22, 2022, in Wuhan, China. (Han Zhilin/VCG via Getty Images)

Charles Clancy, senior vice president at MITRE and GM of MITRE Labs, told Fox News Digital that every major company operates satellites with some level of automation to begin with, including how they manage orbits, schedule data uploads and downloads as well as optimize missions, so this is just another step in the evolution of that process.

“As AI has continued to evolve, it has been able to take over more and more tasks from human operators, allowing humans to focus more on the big picture,” Clancy said. “Sometimes this advanced automation is code that sits on the satellite, and sometimes it is code that sits on the ground and sends instructions to satellites.”

Clancy also stressed that while the specifics of the AI model remain unknown, it does not appear to be a “particularly revolutionary” example, most likely an imagery-based model using an algorithm to pick out ground targets. He pointed to companies like BlackSky who already use similar AI optimization for operations.

In fact, AI will likely make it easier for programming satellites since they only make contact with ground stations “a handful of times” in a typical 90-minute orbit, Clancy noted.


A Chinese rocket is seen blasting off in Xichang, China

A rocket carrying a satellite blasts off from the Xichang Satellite Launch Center on Dec. 29, 2022, in Xichang, China. (Li Jieyi/VCG via Getty Images)

Matt McInnis, a senior fellow for the Institute for the Study of War’s China Program, told Fox News Digital that Beijing views AI as the key tool to help it “leapfrog” the U.S. military for superiority and “allow them to make decisions in potential conflicts much quicker and more accurately.”

“China is investing a lot of in artificial intelligence across the board, but the highest priority for it is how it can help them transform their military into a true world-class power and, frankly, surpass the United States,” McInnis said, adding that it is “a critical component” for that strategy.

McInnis referred to recent revelations by the Israel Defense Forces that it used AI during the 2021 Gaza conflict, which helped the Israelis make fast decisions and also determine likely locations for terrorists, which led to the capture of two enemy combatant leaders.

“Certainly, our concern is keeping up with that in our ability to observe the battle space, observe potential conflict areas, identify targets and then process those to make decisions, and there’s AI that’s going to be part of that additive intensification of targets,” McInnis said. “Then there’s the AI that’s part of decision-making about targets, which is even more complex.”

A Long March 3B carrier rocket

XICHANG, CHINA – December 29: A Long March 3B carrier rocket carrying the Shiyan-10 02 satellite is ready to blast off from the Xichang Satellite Launch Center on December 29, 2022 in Xichang, Liangshan Yi Autonomous Prefecture, Sichuan Province of China.  ((Photo by Li Jieyi/VCG via Getty Images))


“I think that’s really where China wants to go, where they can do it all in the AI-enabled cycle,” he added.

“As China continues to modernize, will they be willing, on the one hand, to give up or to allow artificial intelligence to make decisions that have traditionally resided with humans or humans processing information? Are they going to truly develop AI that they can trust and be able to control?” he said.

“In part, I think there’s a lot of pressure to do this because I think [Chinese President] Xi Jinping still has a lot of doubts about the party, about the loyalty as well as the capabilities of the personnel, and sometimes AI is seen in a way as compensating for the lack of a quality and capability.”


Be known by your own web domain (en)

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *