Of course, it can be argued that the images had no bearing on it, but given that the fake Bloomberg tweet was circulated by big-folk accounts like Russia Today, the country’s national broadcaster, that seems like foolish ignorance.
Fake Reality
Smart fakery is a current problem that has puzzled academics and engineers for years, but is quickly becoming a real-world dilemma in the age of generative AI.
The potential to manipulate markets, foment societal unrest and target individuals for all sorts of horrific online hoaxes is there, and we’re far from prepared to deal with it.
In 2004 (yes, I was a child prodigy, to be honest) I interviewed an expert at the US research firm Gartner named Daryl Plummer about a concept he was already warning about, called “fake reality”, which was technology that was then B. used to create compelling virtual worlds in movies would be used to wreak havoc in the real world.
By this time there were already examples that had occurred. Last year, Piers Morgan was forced to resign as editor of the Daily Mirror after he placed fake photos of British soldiers torturing an Iraqi prisoner on the front page.
Meanwhile, in the run-up to his candidacy in the US presidential election, photos surfaced of Democratic presidential nominee John Kerry speaking with Jane Fonda at an anti-Vietnam rally in 1970.
Veterans called the picture a “slap in the face” and managed to portray Kerry, despite being proven fake, as an anti-military candidate before narrowly losing to George W. Bush.
“Fake reality for sale will be the norm for the next decade, and its growth will depend not on technology but on society’s tendency to embrace the deviant as the norm,” Plummer told me in 2004.
“As people increasingly expect the use of virtual technology rather than real-time human interaction, the psychological outrage to be expected with fake reality will diminish… I suspect there will be some very notable situations where we are deceived and.” make a lot of noise.” will happen. The existing examples are just the tip of the iceberg.”
Although he got the timing wrong, believe it or not, we are now in the age of fake reality.
Voice cloning technology is advanced enough to create unique speech that sounds like the intended person from just a few brief samples of their real audio, and the ability to match this with fake videos will weaponize it.
In the newsroom of a financial newspaper, we must be more diligent than ever before in reviewing incoming anonymous tips. Leaked audio of Qantas and Virgin CEOs discussing a merger might sound like a big reveal, but what if it’s just one? Shareholder with an AI app trying to inflate stocks?
Current Restrictions
The problems of other industries will also be obvious to the people working there.
Voice cloning currently has limitations as it works far more effectively in a short burst of audio and is unconvincing in a live scenario where a human might try to question it, but that’s still a pretty big wiggle room for Fraud.
Earlier this year, AI voice start-up ElevenLabs was forced to issue new terms and conditions after “a rising tide of abuse” as it opened up access to its voice-cloning platform.
Fake speeches by US President Joe Biden are spreading like wildfire, and it’s all hilarious and fun… until, of course, it’s not anymore. (Bad language warning for the end of the video in this tweet.)
Cybersecurity firms are already attempting to sell executive defense services to corporations, but an existential problem is looming in financial markets and in wider society.
Is it the responsibility of market operators to halt stock prices because a tweet goes viral? What does the Commonwealth Bank do when a video of Matt Comyn’s resignation speech goes online when he’s actually on holiday in Byron Bay?
I certainly have no idea, but the worrying thing is that I suspect they don’t know either.
Comments are closed.