Fox Business reported on Monday that Geoffrey Hinton, a Google engineer who is recognized the "godfather of AI," has quit his job, offering this warning: "It is hard to see how you can prevent the bad actors from using it for bad things."
American Family News spoke to Patrick Riley, the CEO of many financial technology startups and a retired Army staff sergeant. Like Hinton, Riley is gravely concerned about the future of artificial intelligence technology – a concern based largely on what those bad actors could do with it.
"While it's easy to imagine the many potential benefits of the emerging AI trend," Riley says, "there's a reason Elon Musk and thousands of others called for a full stop on AI development."
In 2014, long before Musk and others expressed their thoughts, British theoretical physicist Stephen Hawking issued this warning: "The development of full artificial intelligence could spell the end of the human race."
Using AI, anything can be faked
Since the AI chatbot ChatGPT was released in November 2022, "a litany of other AI tools have been launched that will forever change society and technology," Riley explains. At the time of this interview with AFN, he said there were nearly 1,500 AI tools available for public use.
"Some of these tools help you – some do pictures, and others do videos," he describes, adding that more advanced technology can automate tasks, compile code, and more. But despite the positive things AI technology could offer in today's modern world, Riley suggests "[taking] a more measured look at the reality of basic risks by bad actors using these tools."
He recalls a historical radio broadcast that illustrates the reality of risks associated with access to information.
"On October 30, 1938, Orson Welles' Mercury Theatre did a live version of 'War of the Worlds' over the waves, causing mass panic of the entire nation, believing we were truly under threat of a Martian invasion," he notes. "While we would like to think we do not have a gullible population, it really boils down to access to information."
Along those lines, Riley notes that a pair of Russian pranksters continue to be able to dupe their target. Federal Reserve chairman Jerome Powell thought he had spoken to Ukrainian President Volodymyr Zelensky in January, but it was all part of an AI prank. Similar pranks have been performed on Canadian Prime Minister Justin Trudeau, Polish President Andrzej Duda, European Central Bank Chief Christine Lagarde, and singer/songwriter Elton John.
According to Riley, "the ability of anyone to fake anything in a digital medium will occur in a few short years, [and the result will be] overwhelming."
Real world examples
Riley poses the following scenarios:
- "What happens when a commander on the battlefield receives a call from their higher commander, ordering them to put troops into harm's way, only to find out that the live video call was in fact their enemy sending their troops to their death?"
- "How will the public know or be able to verify that a news broadcast shared on the web, showing a well-known anchor speaking on an important subject, is real?"
- "How would a newsroom at a major media company know if a video of incriminating evidence they receive is genuine?"
- "Can the American people be sure that a video, received from the White House, of a president's speech is really of the president and not something that a sly hacker inserted into their feed?"
Finally, Riley questions whether the security of a nation can be assured in an age where digital media cannot be trusted or verified.
"The only way to know in five years if a digital media is legitimate will be if it is on a blockchain," he contends. Blockchain is a technology that allows transparent information sharing within a network – information that cannot be altered without all subsequent blocks being altered as well.
He suggests the White House use its own blockchain; thus, any messages not originating from the blockchain would be considered "illegitimate." But, as Riley points out, "[blockchain technology] is still far from regulatory clarity and mainstream adoption in the public."
What's more, the Army veteran points out that, "the military is often slower to implement new technologies." To that end, he wonders: "Is it only a matter of time until national security becomes a casualty of AI's progress?"
