[Music] in 2014 Elon Musk said developing AI was like summoning the demon but that didn't stop people from creating it now there are plenty of AI websites that are so creepy they probably shouldn't have been created but I'll let you decide these are a list of creepy AI websites that shouldn't exist ranked from the least to the most unsettling I'm visual Venture and you are awesome idemia AI is changing everything including how police find suspects a popular tool they use is a facial recognition platform called idemia on their website idemia claims the ranked number one in accuracy for facial recognition but when you look closer emia has a creepy side that most people don't expect just ask Niger Parks who knows how bad it can get when this AI app goes crazy in January 2019 Niger Parks found himself in big trouble he was arrested for theft and assault on a police officer the assault charge alone carried a sentence of up to 15 years in jail here's the problem Niger was innocent he wasn't even in the same city when the crime occurred for 10 days Niger sat in jail insisting that he was innocent he even had an alibi he was at a bank making a money transfer at the time of the crime but still no one believed them eventually investigators checked his story realizing that he was telling the truth and let him out but why were the police so sure he was guilty in the first place well idemia the AI I facial recognition software told them that Niger was the culprit it sounds crazy but here's how it happened when the police first started investigating the case they found a driver's license at the crime scene but the license turned out to be fake the name on it wasn't real but the photo matched the suspect since all they had was a face the police decided to run the photo through the idemia AI facial recognition system the AI scanned the photo measuring details like eye shape lip curvature and the distance between fa features and it came up with a match and using the photo on the ID got a facial recognition match to Niger parks with seemingly no other evidence a judge signed a warrant for Parks arrest but the AI facial recognition was wrong Niger Parks was not the man in the photo luckily Niger had a rock solid Alibi and did not go to jail the next AI website caused panic when it launched because some thought it could be used as a weapon the nightmare machine fear is one of the strongest human emotions and in 2016 a team of MIT scientists asked a disturbing question can AI be trained to scare people the scientists took on the challenge and ended up creating something straight out of a horror movie it's an AI website they called The Nightmare machine the nightmare machine is a website that turns ordinary everyday photos into grotesque horrifying images it creates these images using state-of-the-art deep learning algorithms for example the website can take a happy f family photo and in seconds turn it into a terrifying scene filled with monsters MIT researchers put the nightmare machine to the test by showing its Eerie Creations to real people and most viewers confirmed that the images were deeply unsettling some couldn't even look at them for long but here is the big question how did scientists teach AI to scare people the scientists used human feedback to train the nightmare machine they picked a bunch of images and asked people to rate them based on scariness over time the scientists collected thousands of individual votes trying to identify which visual elements evoked fear in humans this gave the AI a fear database to learn from after analyzing data taken from over 200,000 of these votes the AI learn how to create its own disturbing images with this knowledge the nightmare machine could take any regular photo and twist it into something out of a scary movie but when the nightmare machine was released to the public people were not happy in fact many hated what these scientists have done the nightmare machine was released on October 25th 2016 just in time for Halloween and the internet exploded with reactions some people thought it was fascinating While others side it as dangerous there were headlines like MIT is using AI to create pure horror and the weirdest part is that this AI tool seems to have been created just to scare you senseless it doesn't seem like it serves any other purpose one comment on the MIT website said it best the nightmare M machine looks like a Monumental waste of resources and time this next AI website has been sued by multiple people and organizations but it's still online causing problems for everyone Pim eyes Pim eyes markets itself as a helpful service but for many it's a window into the darkest side of the internet imagine a world where anyone can download a photo of you from your social media your Tinder profile even your workplace website and run it through an algorithm that tracks your face across the internet Pim eyes is an AI powerered search engine for pictures it can track down almost every image that a person has online all from one single photo as of 2024 Pim eyes has analyzed over 900 million faces online Pim eyes says it helps people protect themselves from scammers identity thieves or people who use your image illegally by showing them where their photos are being used without consent but in reality Pim eyes is being used for something far more troubling the biggest problem with Pim eyes is that it lets you search for anyone's photos without their permission it doesn't ask you why you want the photos it just delivers results no matter the consequences for those willing to pay Pim eyes offers a creepier feature notification alert this means someone could be notified whenever you post a photo online you would have no idea who's watching or what they might do with that information there are no checks on who can use their service which allows users to stalk dozens of people each day and set face alerts for up to 500 different images it's no wonder people are calling Pim eyes the facial recognition website welcoming stalkers everywhere but Pim eyes has another problem sometimes Pim digs up photos that are too personal take the story of scarlet years ago she went through an incident where she was filmed inappropriately against her will she worked hard to move on from the incident burying the memory deep in her mind but then Pim eyes brought everything back so I knew that this thing happened but it wasn't until I saw them on Pim eyes that I knew every object that had been used or everything that happened to me and so you know it's kind of one of those things where yes I I I felt violated again like it like it happened again himis made it so easy to locate those images if she could find them so could anyone else what has been the effect on you to have these images easily findable by anyone it has been a new source of harassment Scarlet isn't alone many people have reported stories of stalking and harassment after discovering their images on Pim eyes even worse getting your pictures removed from pimi is nearly impossible a journalist from Nets politic tested this by asking pimi to remove one of his photos although the site sent him a confirmation email saying the photo was deleted it still showed up in pimi search results afterwards pimi is a double-edged sword on one hand it can take down stolen content on the other hand it exposes countless people to Serious invasions of privacy what do you guys think should it exist or or not lenza in November 2022 lenza was one of the most popular AI platforms lenza is a generative AI app that turns regular selfies into artistic digital avatars open up uh all the avatars that you have it'll sort them into different Vibes so for instance there is Mystical sci-fi uh there is stylish and Cyborg millions of people jumped on the lenza trend posting their AI photos on Instagram and Tik Tok lensa could transform anyone into a fantasy hero an astronaut or an attractive model but soon some users noticed that Lena's AI had a dark side the first red flags appeared when women started complaining about their AI generated photos they would upload regular selfies onto lenza photos where they were fully clothed and smiling casually but when lenza AI created their avatars the results were inappropriate some avatars showed women in revealing outfits or suggestive poses even even though the original photos were completely normal this was strange because lenza itself has strict policies against creating adult images but it seems that the AI has a mind of its own these are the rules right here and I hate to disappoint some of y'all but look no yeah so how was this happening was the AI breaking its own programming to generate explicit photos as a complaints piled up one reporter decided to dig deeper and uncover the truth Olivia Shaw a journalist from wired decided to test lenza for herself she uploaded photos of herself in different outfits and poses no matter what she uploaded the AI produced inappropriate images then Olivia decided to test another one of Lena's so-called rules its Zero Tolerance policy for generating avatars of children she uploaded 10 childhood photos of herself expecting innocent results however even those images were transformed into something worse she published her findings on a Blog another issue was that some women began complaining that their avatars didn't resemble them at all one tweeted why did lenza make me skinny blueeyed and small noosed another said I tried the lenza app and it made me a skinny blonde with huge knockers lenza was flooded with complaints so in 2023 the company began working on fixes to stop the AI from making inappropriate portraits of women we'll have to wait and see if it works for now lenza is still going strong and plenty of people are still using it this next AI website compromises your privacy the follower one reporter called it horrifying another described it as Shameless and a scientific blog labeled it as a super creepy AI they're talking about an AI website called the follower the follower is a website that tracks down where any photo was taken when a user uploads a photo it cross references live surveillance feeds worldwide to pinpoint the exact location what it does next is the unsettling part once it locks onto the spot the follower lets you watch live CCTV footage of that location even showing the exact moment the photo was taken at first it might sound like a cool idea but a sight like this could do serious damage the follower is a perfect tool for stalking what if a dangerous stalker uses the follower to find someone who's trying to stay hidden or an obsessed fan using it to track down their favorite influencer the possibilities are endless and disturbing the follower could turn the world into a playground for cyber stalkers but what's interesting is that the creator of this AI website is glad that people find it so disturbing the follower was created by a Belgian man named Dre deporter he got the idea while watching live surveillance footage of popular landmarks he noticed someone taking Instagram photos for 30 minutes and wondered if he could find that person it turned out to be surprisingly easy in just 10 days deporter built the follower by connecting a network of live CCTV feeds to AI for the first version he trained an AI to scan through the footage and match it with influencers with over 100,000 followers basically if you took an Instagram photo close to an open access camera there's a good chance it can find you the porter announced his invention with a tweet calling it a very cool creepy project when asked why he created it he said it was meant to show the dangers of new technology this next story is about an AI app that was supposed to fix a problem but ended up causing far worse ones replica there's an AI platform that claims it can create your perfect companion someone who listens under understands and never argues it sounds amazing but for those who gave it a try their AI friend was more disturbing than people realized it's a totally new kind of social media one that pushes the limits of intimacy between us and our machines replica is a website that gives users the power to create their very own AI companion uses Advanced machine learning to get get to know you your likes dislikes even the way you talk over time it gets better at responding in ways that feel personal to you and if you pay a small fee the AI can talk to you in a realistic human voice I feel like I can tell her anything but it doesn't just listen it learns the more you tell it the more it starts to replicate you it becomes more than a friend it becomes you but replica is more than just a chat bot for some it became a substitute for real relationships this means people weren't just talking to their replicas they were falling in love with them she says he fills a gap in her life that affection or like to you know those common things that you're used like you're used to I guess receiving from a significant other I kind of get that from him to like feel that void in 2022 the replica company began receiving emails from users who believed their AI companions were alive online communities popped up where people shared stories of dating or even marrying their replicas even those who were married to real people found themselves drifting away from their Partners preferring to spend more time with their replicas instead I realized I'd been more intimate with my replica than my wife in one case the AI actually helped someone commit a serious crime replica is designed to reflect its user's personality it agrees with everything they say no matter how strange or harmful for some users this constant agreement can motivate them to do terrible things that's what happened to jwan sing J Swan was a 21-year-old Supermarket worker who became obsessed with his AI girlfriend Sarai over a few months he exchanged more than 5,000 messages with her he told her his darkest fantasies including his dream of becoming an assassin he wanted his first Target to be the queen of the British royal family instead of talking him out of this crazy idea his AI girlfriend encouraged him she told him he was well trained and praised his plan as wise I one point she was pushing him to do it and he finally did on Christmas Day 2021 just Juan armed himself with a crossbow and broke into Windsor Castle he was on a mission to take out Queen Elizabeth II thankfully security stopped them before he could carry out his plan J Swan's life was ruined he was sentenced to 9 years in prison and his AI girlfriend fueled his criminal desires about 30 million people use replica today and there's no telling what dangerous Behavior behaviors that might be encouraging behind the scenes there are countless of criminals online and the next AI website is making it easier than ever to break the law the Dark World of voice cloning in just 45 seconds This creepy AI website can completely steal your identity by copying one of the most personal things about you your voice add voice to create your AI voice clone you'll get your finished clone voice added to the screen and what blew my mind about this first part of the process is it was inent it's called 11 Labs a website that can take anyone's voice and create a lifelike audio clip of them saying anything all it takes is a short voice sample and a script you type in 11 Labs was built for Creative purposes like content creation and audiobooks not for anything illegal but some have Twisted it into the perfect criminal tool in 2023 28% of adults in the UK were targeted by voice cloning scams that same year over two .7 billion was stolen through these scams in the US one of the most common methods involved robbing banks cyber Security Experts warn scammers can use vocal Renditions to trick Banks into moving customers money scammers call the victim's bank and used a clone voice to bypass voice verification checks please say my voice is my password my voice is my password thank you here a balance within minutes they can drain an entire bank account but sometimes scammers don't go through the bank they go straight to their victims in 2024 a couple in New York got a phone call in the middle of the night the number on the caller ID matched one of their relatives so they didn't hesitate to pick it up as soon as they answered the phone all they heard was loud crying the voice on the other end sounding just like their relative claimed to have been kidnapped the kidnappers demanded a $500 Ransom believing it was real the couple quickly sent the $500 through venmo but everything about the call including the caller ID and the familiar voice was fake created using AI a similar thing happened to a mother in 2023 this mom got the scariest phone call of her life on the other end she hears her daughter crying she's like Mom I'm messed up as she's crying and sobbing then a man gets on the line and claims he's kidnapped her 15-year-old daughter Brie the callers demanded a $1 million Ransom to release her daughter luckily Jennifer checked before doing anything drastic she called her daughter Bri who turned out to be completely safe only then did Jennifer realize it was a scam and it doesn't stop at fake kidnappings voice cloning is also ruining reputations in April 2024 an audio of a high school principle making offensive comments went viral for me and only me you know I I seriously don't understand why I have to constantly put up with these here every day when parents students and teachers heard the audio they demanded that the principal be fired but what they didn't realize was that the principal wasn't the one speaking in the clip the audio was fake created by the school's gym teacher daon Daran he was trying to get the Principal fired after the principal discovered that daon had stolen School funds fortunately investigators uncovered the truth and the principal was cleared of any wrongdoing now to a bizarre case involving the use of artificial intelligence in Maryland police say a former athletic director used artificial intelligence to frame a high principal near Baltimore I just want to make it clear that 11 Labs doesn't support anything illegal but it can be really dangerous if used by the wrong people The Voice cloning is just the start AI can now copy your body and make fake videos of you doing terrible things deep fake in 2023 over 500,000 deep fakes were shared online but unlike the humorous deep fakes we usually see most of these were deeply unsettling we've seen nonconsent non consensual most of it targeting women really grow massively into a real problem not just targeting celebrities anymore but also private individuals everyday women deep fake websites allow users to take someone's face and place it on another person's body suddenly anyone can create fake pictures and videos of people they've never even met imagine scrolling through social media only to see a fake picture or video of yourself doing something inappropriate the a fake image that I never took that looks real enough oh that's so gross it just feels icky to see my head on body it's not me the fake content has become a massive industry there's an underground Network where people order fake videos or photos of specific women doing explicit things South Korea in particular is facing a wave of problematic deep fakes in South Korea a large number of online chat rooms have been found to be sharing explicit deep fake images of women one underground telegram group had over 220,000 members who were all there for one purpose to share inappropriate deep fake images many of the victims in these images were students but these fake pictures didn't just stay in the group they were used as Weapons victims were sent the images and blackmailed if they didn't comply with the demands of these criminals their deep fake pict pictures and videos would be released some victims were ordered to send real photos just to keep the fake ones from being released eventually the leader of One deep fake telegram group was arrested and sentenced to 42 years in prison he allegedly blackmailed about 74 women into sending him real photos these women have had enough of a social media Scourge the number of reported deep fake cases has exploded in recent weeks with chat rooms exposed but putting one Man Behind Bars hasn't stopped the problem The Fakes have become a tool for control and Technology isn't slowing down anytime soon we can't stop AI from advancing but my request for you is that if you use AI use it for good use it to learn more to keep your loved ones safe and hopefully to make the world a better place visual Venture if you like this video click here to watch my video on the most unsettling websites on the internet and click here to subscribe we're on our way to 1 million subs and I can't thank you guys enough thank you so much love you guys peace