AI-generated images are getting harder and harder to find. There is a common sayingOf course, but it’s telling that with every new AI image model the minimization is done. You can trust AI-generated hands with too many or too few fingers on people, for example; Nowadays, it is not necessary.
One particular area of weakness for AI image models is text generation. The image itself may seem convincing, but look closely at the words, and you’ll often find they’re not indeed Right maybe some are accurate, while others are pretty close, but, in many cases, you’ll see a lot of inconsistencies: lots of repeated letters, letters that aren’t really letters, letters that blend in and blend into each other. A lot of these things remind me of how The language of Star Wars appearsAt least when the AI is trying to mimic American English.
OpenAI’s Images 2.0 is capable of generating highly realistic AI images
But the latest AI models are getting better with text generation. In fact, OpenAI’s latest model for ChatGPT, Images 2.0, Can render extremely realistic textAnd A a lot From that – to the point where I’m not sure many of us (or anyone of us) will find it. According to OpenAI, Images 2.0 is the company’s first image model with thinking capabilities: that means the model can take its time breaking down each step of the request, generating more detailed or accurate images, as well as the ability to generate up to eight images from a single prompt (although this is only available to paid subscribers). Free users can still take advantage of Images 2.0 benefits, such as how it searches the web for information and double-checks its work. The company says that “results look less AI-generated and more deliberately designed,” which essentially means that images generated using ChatGPT will be harder to spot going forward.
The company is very confident about this latest model. It shows the number of different types of images it can create – not just photorealistic pictures, but computer UIs, magazine collages, screenshots of rice paddies (it is A lot to generate), a magazine page and a handwritten essay. They also mean handwriting down to coffee stains on paper. You can scroll through OpenAI’s official announcement post to see these examples How shockingly real they are. The post includes other examples, such as highly realistic photographs, graphic novel pages, movie posters, and images with different aspect ratios – up to the iPhone’s panorama view.
What do you think so far?
All of these developments are as impressive as they are painful, but for me, it’s the improvements to the text that really take it to another level. Many of the models are doing a pretty good job of generating images that trick users into thinking they’re real, but the level of detail in the text and writing on these examples is something I’ve never seen before. I asked ChatGPT to generate me a menu for an Italian restaurant, highlighting five dinner courses and two desserts—I left the exact dishes to the AI’s discretion. As far as I could tell it managed to generate something realistic with dinner and dessert entrees without errors. Then I asked him to generate a newspaper entry that the Red Sox and Yankees would switch cities. He did that too, again with no glaring mistakes.
Credit: Lifehacker
I am not saying that these are pictures perfect: They still have an AI “glow”, which a trained eye or close observer will be able to notice. OpenAI says that Images 2.0 struggles with certain complex tasks, such as puzzles, as well as details found on hidden or oddly placed areas, such as contrasting surfaces. But none of that really matters when the images this model creates are impressive enough to fool most people who pass by. Infographics, photographs, maps, comics, movie posters, you name it: people are going to use this tool, and you’re going to start seeing a lot more AI images in your life—often without even realizing it.
