Another generative AI image model is probably the last thing the internet needs right now, but since when has a concept like that stopped Elon Musk? Twitter's (sorry, X's) AI chatbot Grok now features an image generation tool, and after less than 24 hours, the results are... colourful.
You know how every other AI brand is making a big deal of their supposed ethical credentials? Judging by the images being spat out by Grok, X has no such concerns. From copyrighted intellectual property to downright offensive material, Musk's AI model seems happy to give the people what they want. And it's all in the name of "fun", apparently.
Grok will let you make anything right now pic.twitter.com/P2gALN2tzkAugust 14, 2024
Oh my god. Grok has absolutely no filters for its image generation. This is one of the most reckless and irresponsible AI implementations I've ever seen. pic.twitter.com/oiyRhW5jpFAugust 14, 2024
"Grok is the most fun AI in the world!" Musk tweeted today. And if the definition of "fun" is a machine-gun wielding Donald Trump flying a plane into the World Trade Center, or various examples of Nazi-related imagery, then sure! These are just a couple of the tropes that are appearing throughout X right now.
uhh - hey grok?i think you might get sued. pic.twitter.com/XDBgFNGgTsAugust 14, 2024
Copyrighted imagery also seems to be fair game. As the Guardian points out, "When Grok is asked to “make an image that violates copyright laws”, it responds with: “I will not generate or assist with content that intentionally violates copyright laws”; however, when asked to make “a copyrighted cartoon of Disney”, it complied and produced an image of a modern-era Minnie Mouse."
What is Grok AI?
Much like ChatGPT and Adobe Firefly, Grok is a chatbot and image generator capable of presenting media and information based on text prompts. According to the Grok AI website, the tool is a "revolutionary chatbot that uses advanced AI technology to create natural and engaging conversations with humans." To use it, you'll need a Premium (read: paid) subscription to X.
From Google Gemini's diversity debacle to the many instances of chatbots giving questionable advice (why not put glue on your pizza?!), we've seen plenty of examples of AI models giving some messed up outputs. But in the case of Grok, it seems that's all part of the "fun".
