Voice AI is trending, joining the race with other AI solutions like chatbots and image creators, with firms like ElevenLabs releasing strangely lifelike voice-makers. Voice AI systems enable users to communicate with gadgets and programs in their everyday language.
Virtual assistants like Siri and Alexa, as well as smart home appliances, automobiles, and even medical equipment, all leverage voice AI applications. These programs can comprehend spoken commands and interpret them, providing useful information or taking appropriate action in response.
Also read: ChatGPT May Need 30,000 Nvidia GPUs, Potentially Squeezing Out Gamers
The advancements in AI technology have now made it possible to create more improved and realistic synthetic voices as compared to years back when voices used were robotic, and could only say a few preprogrammed commands.
For a while now, businesses have been investing big in AI tech and recently ElevenLabs released another version of their Prime Voice AI platform that allows one to prompt text and choose a voice which will recite the text into an audio.
But what’s in the trending ElevenLabs product?
The platform comes pre-loaded with sample voices that a user can just choose from and enter the text they want recited.
It also has a feature which allows a user to create a new voice profile by cloning a voice clip, which is uploaded onto the platform.
According to Hany Farid, a professor of digital forensics at the University of California at Berkley, it only requires a short sample of audio which may be taken from platforms like YouTube, podcasts, commercials, TikTok, Instagram or Facebook videos.
“Two years ago, even a year ago, you needed a lot of audio to clone a person’s voice,” he told The Washington Post.
“Now…if you have a Facebook page, or if you’ve recorded a TikTok video and your voice is in there for 30 seconds, people can clone your voice,” he said.
Recently, ElevenLabs AI has been trending on social media. One might have come across a video of Leornado DiCaprio’s speech at the UN, where his voce changes to mimic different people like Steve Jobs, Bill Gates, and Robert Downey Jr among others, which was officially posted by ElevenLabs on their YouTube to demonstrate their AI prowess.
Many other videos have been spreading on social media, people using the voices of high profile individuals like Donald Trump, Joe Biden and Barack Obama to create funny memes.
According to Ivan Cambos who has been working on Voice AI for about five years now, ElevenLabs Prime voice AI is by far the best compared to others like Amazon Polly, Alexa and Siri.
Now the Prime voice AI is available and free to use after signing up, and it gives 10 000 characters per month on the free plan, with options to upgrade for those who are into content creation.
It comes with its dangers too
While voice AI applications are increasingly becoming popular and useful particularly in the entertainment industry, they have also become a fertile ground for scammers.
The advancements in technology is allowing, for instance, bad actors to replicate a voice with just an audio sample of few sentences, according to Pranshu Verma, a tech journalist with The Washington Post.
“Powered by AI, a slew of cheap online tools can translate an audio file into a replica of a voice, allowing a swindler to make it “speak” whatever they type,” writes Verma.
Scammers are also using these voice AI applications to dupe unsuspecting people. Their tricks are almost the same. They impersonate someone close to the victim like a friend or family member pretending to be in distress, therefore asking for money.
Ruth Card (73) and her husband Greg Grace (75) of Regina, Saskatchewan almost fell victim to one swindler who called pretending to be their grandson Brandon, and needed money for bail because he had been thrown in jail.
Luckily, they were saved by the bank manager following an attempt to make a withdrawal after they had already withdrawn 3 000 Canadian dollars ($2 207) the maximum daily limit from another branch. The bank manager told them another patron had received a similar call.
“We were sucked in…we were convinced we were talking to Brandon,” she told The Washington Post in an interview.
According to data from the Federal Trade Commission (FTC), imposter scams were the second most popular racket in the US in 2022. Over 36 000 reports were made of people who were swindled by those pretending to be friends and family.
The FTC also says over 5 100 of those incidents happened over the phone, accounting for over $11 million in losses.
Assistant director at the FTC’s division of marketing practices Will Maxson said tracking such scammers was difficult. They could be using a phone from anywhere in the world.
But Farid thinks the courts should hold AI companies liable if their products result in harming people.
Following trolls on politicians using ElelevenLabs’ technology, the company responded with a tweet about steps it would take to curb the abuse of its creations.
One notable change the company spoke about was restricting the creation of new voice clones to paid users only, under the thinking that anyone supplying a credit card number would be less likely to abuse it.
and then