wp-plugin-hostgator domain was triggered too early. This is usually an indicator for some code in the plugin or theme running too early. Translations should be loaded at the init action or later. Please see Debugging in WordPress for more information. (This message was added in version 6.7.0.) in /home4/scienrds/scienceandnerds/wp-includes/functions.php on line 6114ol-scrapes domain was triggered too early. This is usually an indicator for some code in the plugin or theme running too early. Translations should be loaded at the init action or later. Please see Debugging in WordPress for more information. (This message was added in version 6.7.0.) in /home4/scienrds/scienceandnerds/wp-includes/functions.php on line 6114Source:https:\/\/techcrunch.com\/2023\/03\/29\/ai-pope-midjourney-goncharov\/<\/a><\/br> Last fall, Tumblr users raved about Martin Scorsese\u2019s 1973 film \u201cGoncharov<\/a>,\u201d starring Robert De Niro and Al Pacino. They hailed the lesser-known Scorsese classic as the greatest mafia movie ever made. It was ahead of its time, never earning the acclaim it deserved.<\/p>\n By now, we should all know that \u201cGoncharov\u201d is not a real film, even though Scorsese himself got in on the joke<\/a>. It\u2019s just a Tumblr<\/a> meme that spiraled into a site-wide gag, as users toiled away in shared Google Docs, collaboratively working together to \u201cyes, and\u201d one another with such sincerity and precision that the made-up movie could seem legit.<\/p>\n As AI image generation gets better, the internet feels a bit like Tumblr in the moment of \u201cGoncharov.\u201d We are creating new collective realities, only this time, we are accidentally generating them simply by posting without much thought.<\/p>\n When AI image generator Midjourney opened access to the Midjourney 5<\/a> model two weeks ago, its hyper-realistic outputs quickly went viral. Some of the resulting images illustrated the dangers of these easily accessible tools, like a series of fake photos depicting Donald Trump\u2019s arrest<\/a>.<\/p>\n The dramatic set of images shows Trump forcibly resisting arrest, then sprinting away from a squad of police officers, among other scenes. The images, which spread like wildfire around the internet, were generated by Eliot Higgins, founder of investigative journalism outlet Bellingcat. Higgins\u2019 initial tweets made it clear he was just playing around with AI image generators, but they quickly spread without the accompanying context.<\/p>\n Other viral AI images are more benign, like a widely circulated photo of Pope Francis II serving a high fashion look<\/a> in a stunning, white Balenciaga coat. Just one of the viral tweets<\/a> about the image was viewed over 26 million times.<\/p>\n That wasn\u2019t even the only Balenciaga-related AI meme to go viral in the last week. A creator using the name demonflyingfox<\/a> has been using Midjourney images, Eleven Labs speech and D-ID animation to make surreal, AI-powered parody videos. His most widely spread video is \u201cHarry Potter by Balenciaga<\/a>,\u201d a high-fashion-style advertisement in which synthetic renditions of characters from the franchise say things like, \u201cYou are Balenciaga, Harry.\u201d<\/p>\n Demonflyingfox told TechCrunch that he intentionally tries to make his videos weird enough that no one could accidentally believe that they\u2019re real. (It doesn\u2019t take a savvy viewer to know that Joe Rogan has never actually interviewed Jesus Christ<\/a> and Dumbledore regrettably did not wear a leather jacket and sunglasses in the Harry Potter movies.) He also deliberately uses Midjourney 4 instead of the updated software so that his characters don\u2019t look so realistic that they could come off as believable.<\/p>\n \u201cWhat I\u2019m doing is so obviously fake that I don\u2019t really worry about spreading misinformation, nor is it my intention,\u201d demonflyingfox said. \u201cBut I know of the power of the tools and that it\u2019s very easy right now to do so.\u201d<\/p>\n The more believable the premise, the more likely we are to register a synthetic image as fact. If you\u2019re not an expert on papal attire, you could feasibly believe that the Pope has a kick-ass white puffer jacket for winter outings \u2014 this is the same Pope whose prog rock album<\/a> got reviewed on Pitchfork, and who once worked as a nightclub bouncer<\/a> in Argentina.<\/p>\n Even supermodel Chrissy Teigen fell for the joke. \u201cI thought the pope\u2019s puffer jacket was real and didn\u2019t give it a second thought,\u201d she tweeted<\/a>. \u201cNo way am I surviving the future of technology.\u201d<\/p>\n The Trump AI arrest images also went viral, but could be fact-checked quickly enough by anyone willing to scan their Twitter timeline for evidence of the real news breaking. That process is a little less obvious for AI-generated historical events.<\/p>\n Just days ago, Reddit user u\/Arctic_Chilean posted a set of AI-generated photos<\/a> on the Midjourney subreddit, claiming that they illustrated \u201cThe 2001 Great Cascadia 9.1 Earthquake & Tsunami,\u201d which hit the U.S. and Canada. Within the confines of that subreddit, which is dedicated to experiments with Midjourney\u2019s AI tools, it\u2019s obvious that the Great Cascadia Earthquake is not a real event. But if r\/Midjourney is just one of many subreddits on your feed, it\u2019s easy to consume the images without giving them a second thought.<\/p>\n \u201cWas I the only one who was like \u2018How come I don\u2019t remember this happening?\u2019\u2026 until seeing the subreddit?\u201d one Reddit user wrote<\/a>.<\/p>\n The images parallel reality closely enough that they could be true. The Cascadia subduction zone is a real fault line near the Pacific Northwest and the site of a massive, devastating earthquake in 1700. People in the region worry that a disaster on that scale<\/a> could occur again in our lifetime, a knowledge that imbues the AI-generated scenes of potential future events yet to unfold with a sense of foreboding.<\/p>\n Sometimes, AI tools like ChatGPT and Bing AI \u201challucinate\u201d \u2014 a term used when they confidently answer questions with fake information. Awash in a sea of synthetic imagery, we might all be on the cusp of collectively hallucinating.<\/p>\n \u201cI was about to trash you for posting old news in this subreddit lol,\u201d a Redditor in r\/midjourney commented<\/a>. \u201cThese look so real it\u2019s insane.\u201d<\/p>\n Internet communities have a tendency to converge around a niche idea and build out in-depth lore about it, like an act of collaborative world-building (see: \u201cGoncharov\u201d). Naturally, the same thing happened with these fake historical events, as Reddit users began creating their own history about how the 2001 earthquake impacted global politics.<\/p>\n \u201cDespite the magnitude and devastation caused by the Great Concordia Earthquake of 2001, it seems that 10 years later, few people remember the event,\u201d wrote another user<\/a> in a made-up news article, misnaming the earthquake that never happened. \u201cThis is likely due to a combination of factors, including the global war on terrorism, the Iraq War, and the Great Recession.\u201d<\/p>\n The Redditor who originally posted the images theorized about how a major disaster just before 9\/11 would impact the War on Terror, positing<\/a> that the calamity\u2019s economic impact might undermine support for the invasion of Iraq.<\/p>\n As widely available AI image generators rapidly become more sophisticated, their creations might outpace our ability to adjust to a flood of believable but completely false images. For better or worse, it\u2019s now possible to create our own \u201cGoncharov\u201d in an instant, turning any nebulous fiction into something tangible and putting it online.<\/p>\n While many of these creations are harmless, it\u2019s not difficult to imagine how synthetic images might manipulate public knowledge of current or historical events.<\/p>\n It turns out that \u201cBalenciaga Pope\u201d was the brainchild of a 31-year-old on shrooms<\/a>, who told BuzzFeed News that he hadn\u2019t considered the consequences of AI images. \u201cIt\u2019s definitely going to get serious if they don\u2019t start implementing laws to regulate it,\u201d he said.<\/p>\n
\nFrom Balenciaga Pope to the Great Cascadia Earthquake, AI images are creating a new reality<\/br>
\n2023-03-29 21:59:38<\/br><\/p>\n