An unidentified individual on Discord has shared alleged leaks of OpenAI’s upcoming upgrade to its AI image generator, causing a stir among AI enthusiasts worldwide. According to a report by The Decoder, the new version, known as DALL-E 3, offers significant improvements in visuals as well as understanding user prompts and generating image texts compared to its predecessor. The rapid and impactful upgrades by tech companies in the AI field are truly impressive, and OpenAI is poised to once again revolutionize the industry with its upgraded AI image generator. As more people incorporate AI tools into their daily lives, the impact of DALL-E 3 will become evident. In this article, we will delve into the alleged improvements of DALL-E 3 and compare it to competing AI image generators to showcase its technological advancement.
DALL-E, OpenAI’s proprietary AI image generator, is not just known for its world-renowned ChatGPT text generation tool but also for its ability to produce images based on user prompts. Currently, the latest version is DALL-E 2. This version, similar to its competitors, offers enhanced visuals compared to its predecessor, DALL-E 1. However, OpenAI seems poised to take a huge leap forward with the upcoming release of DALL-E 3. The Decoder reported that a leaker shared information about the new AI image model on a Discord channel back in May. The leaker, claiming to be part of an alpha test for OpenAI, shared several samples of DALL-E 3 to substantiate their claims and generated a lot of online attention.
One of the exciting aspects of DALL-E 3 is its ability to create higher-quality images compared to its predecessor. It addresses specific challenges encountered by AI models, such as accurately rendering hands, fingers, and toes. Previous models often displayed incorrect finger counts or merged fingers. Additionally, DALL-E 3 has made significant improvements in understanding user prompts, further enhancing its capabilities compared to DALL-E 2.
To fully comprehend the significant upgrades in DALL-E 3, it is essential to compare images created by the second and third versions. Let’s analyze images generated from the prompt “A painting of a pink jester giving a high five to a panda while in a cycling competition. The bikes are made of cheese, and the ground is very muddy. They are driving in a foggy forest. The panda is angry.” The top image represents DALL-E 2, while the bottom image represents DALL-E 3. The DALL-E 2 image depicts jester and panda hands touching while riding cheese-colored bikes. However, the hands appear distorted and fused together. On the other hand, the DALL-E 3 image accurately portrays the panda’s anger, the muddy path, and an additional cyclist, creating a more authentic biking competition scene. The bikes even give the impression of being made of cheese. Most notably, DALL-E 3 effectively captures the high-five gesture between the jester and the panda.
Let’s also consider the same prompt interpreted by another AI image generator called Midjourney. In Midjourney’s interpretation, we see a concrete path, a happy bear, but no jesters or cheese bicycles. It is evident that Midjourney struggles to match the quality delivered by DALL-E 3. Furthermore, DALL-E 3 possesses the unique ability to incorporate text into its generated images. For instance, when given the command “An image of an angel holding the sun and moon. Above the angel, it says ‘BE NOT AFRIAD.’ In the background, the entire universe is visible. The image should have a fantasy art vibe, 8k resolution, and evoke beauty and emotions,” DALL-E 3 successfully recognizes the grammatical errors intentionally included by the leaker and produces an image of an angel with the words “Be not afraid” above it.
Another impressive feature of DALL-E 3 is its ability to avoid concept spillover or mixing different content concepts. For example, when given the prompt “A group of farm animals (cows, sheep, and pigs) made out of cheese and ham on a wooden board. There is a dog in the background eyeing the board hungrily,” DALL-E 3 accurately depicts the farm animals made of cheese and ham on a wooden board while ensuring that the dog in the background remains realistic. On the other hand, tools like Stable Diffusion and Midjourney tend to turn every animal in the image into food models, failing to maintain the realism of the background dog.
In conclusion, leaked images allegedly created by OpenAI’s upcoming AI image generator, DALL-E 3, showcase significant improvements compared to its predecessor as well as competing AI image generators. DALL-E 3 demonstrates better understanding of user prompts and delivers higher-quality images. However, one notable advantage of other tools is their availability to the public. As of now, OpenAI has not confirmed the development of DALL-E 3 or a similar upgrade. For more digital tips and trends, visit Inquirer Tech. Your subscription has been successful. Don’t miss out on the latest news and information. Subscribe to INQUIRER PLUS to gain access to The Philippine Daily Inquirer and over 70 other titles, share up to 5 devices, listen to the news, download articles as early as 4am, and share articles on social media. Contact 896 6000 for assistance.
Denial of responsibility! VigourTimes is an automatic aggregator of Global media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, and all materials to their authors. For any complaint, please reach us at – [email protected]. We will take necessary action within 24 hours.
