Deepfake technology uses artificial intelligence and deep learning to generate fake images or videos. This technology can seamlessly replace faces in photos or videos, allowing you to swap one person’s face onto another person’s body, simulating Hollywood-style special effects.
Whilst it may be used to occasionally create funny videos and images on social media that are mostly harmless, the are some worrying implications attached to this. Both politicians and celebrities have been victims of deepfakes, as has the Queen. In 2019, the character of Jon Snow issued an apology for the (dismal) ending of Game of Thrones – perhaps one of the more innocent examples of deepfake technology.
Deepfake video editing programmes are no more complicated than traditional video editing software, meaning this type of editing is becoming more accessible and more frequently used.
As technology improves, deepfakes are also becoming harder to spot. In 2018, US researchers discovered that deepfake faces don’t normally blink, but as soon as this weakness was identified, deepfakes with blinking began to appear.
Perhaps the most concerning impact of deepfakes is that they could create a zero-trust society, where people are unable to distinguish truth from falsity – and we’ve already had glimpses of the sort of trouble and deeply entrenched societal issues that this can dig up.
From smart speakers to smart TVs and doorbells, many of our homes are now inundated with household companions that respond to voice commands.
Experts have warned that these smart devices could be more invasive than we think. In a dystopian-style future – which some people believe we’re heading towards – these devices could be used as surveillance equipment for massive information collection that could be sold to third parties.
Some studies have found that digital assistants can be ‘awake’ and listening even when not activated by voice commands. In fact, the devices are actually listening whenever they are turned on, meaning they could potentially build profiles on anyone in the room, monitoring everything from conversations to daily habits, and use this for relevant advertising.
Some smart TVs are equipped with microphones that are always on, meaning they have the potential record conversations when you aren’t watching TV.
Something to consider when you’re searching for your next smart device!
We may all be used to multi-factor authentication and fingerprint or facial recognition when it comes to verifying our identity, but you might not be quite so familiar with voiceprint recognition.
In the same way that everyone has a unique fingerprint, we all have a unique voiceprint that can be used to identify us, and recognition systems can quickly identify a person after listening to just 8 to 10 words.
A budding use of this tech is the improvement of customer service, where customers can rely on their voices instead of having to remember passcodes or answering security questions to access their accounts.
Whilst this may sound ideal for those of who can never remember their security questions, you might not be so keen on the idea, however, when you consider the possibility of voice cloning…
Criminals could potentially steal your voice by recording a short sample of your speech and cloning it, using it to access your personal information, or even speak to your employer or family members. And whilst it won’t take you too long to notice if your credit card is stolen, identifying voice theft is a different matter entirely.
In recent years, robots have taken over many repetitive, boring jobs that humans have been happy to give up. However, concerns have always persisted that robots will soon be capable of increasingly complex tasks and will render a vast proportion of the human workforce obsolete.
You would expect that writing would be one of the jobs that is most safe from a Terminator-style robot takeover – especially when you take into account tone, style, and voice – but, unfortunately, this is no longer the case.
Artificial intelligence research lab, OpenAI, made a text generator so successful at producing text that it’s capable of writing a basic news article with no input from a human editor.
The creators of this tool didn’t release the dataset used to train the text generator due to concerns that it could be misused – which is perhaps the most concerning aspect of the story.
If we could soon see robots writing our news stories, who knows what the tasks they will be capable of next?
We hope you enjoy this spooky season and, if you experience any IT scares, please get in touch with our team, who will be happy to help.
Talk through your IT needs with one of our friendly technicians – no commitment, no cost!
Complete our contact form and one of our experts will be in touch, or call our team on 01332 374444.
Array ( )