Microsoft’s FREE Bing AI Art Generator vs Midjourney V5 Direct Prompt Comparison

MattVidPro AI
21 Mar 202331:34

TLDRIn this video, the host explores the new Dolly 2 algorithm incorporated into Microsoft's Bing AI Art Generator, comparing its image generation capabilities with Midjourney V5. The host expresses excitement about the potential of the Dolly 2 algorithm, noting its detailed and creative image outputs, which are competitive with Midjourney in many aspects. After encountering initial difficulties in generating images through Bing's chat feature, the host successfully generates a crocodile image and is impressed by the quality. The video continues with a series of image generation tests with varying prompts, comparing the outputs of Bing's Dolly 2, the original Dolly, and Midjourney V5. The host concludes that while the new Dolly 2 has significantly improved and offers free, unlimited image generation, it still falls short of Midjourney V5's quality. However, the host acknowledges that for users seeking a free alternative, the new Dolly 2 is a substantial upgrade from its predecessor and provides value in its current state.

Takeaways

  • 🤖 Microsoft has integrated OpenAI's DALL-E algorithm into Bing as a new feature called Bing Creator, offering AI-generated images through the Bing chat feature and Microsoft Edge.
  • 🎨 The new DALL-E algorithm, referred to as DALL-E 2, has shown significant improvements in image quality and detail, making it competitive with Midjourney in various aspects.
  • 🔍 Users can access Bing Creator by using Microsoft's new Bing chat feature or directly through Microsoft Edge, marking the first browser with an integrated AI-powered image generator.
  • 📈 The updated DALL-E model within Bing has demonstrated better performance in generating images that are more coherent, detailed, and realistic compared to the previous version.
  • 🚀 Despite initial difficulties in getting Bing to generate images through chat, the separate website for image generation proved effective, showcasing the algorithm's capabilities.
  • 🤔 The human brain processes visual information much faster than text, which is why Bing is focusing on creating visual tools to enhance the search for information.
  • 🌐 Bing Image Creator is currently only supported in English and is accessible to all Bing users, indicating a strategic move towards integrating more visual content into search experiences.
  • 🎭 The new DALL-E model has shown the ability to generate complex and creative images, such as anthropomorphic characters and specific scenarios, which were difficult for the older model.
  • 📊 When compared directly to Midjourney V5, the new DALL-E model produced images that, while improved, still did not match the level of detail and realism of Midjourney's output.
  • 💡 The new DALL-E model's integration into Bing represents a significant step forward in AI-generated content, offering users a free and accessible platform for creating visual content.
  • 🌟 Midjourney V5 continues to excel in generating highly realistic and detailed images, setting a high benchmark for other AI image generation models to match.

Q & A

  • What is the new feature Microsoft has decided to incorporate into Bing?

    -Microsoft has decided to incorporate OpenAI's DALL-E algorithm into Bing, specifically through a feature called Bing Creator, which allows users to generate images using the new DALL-E 2 algorithm.

  • How does the new DALL-E 2 algorithm compare to Midjourney in terms of image generation?

    -The new DALL-E 2 algorithm appears to be competitive with Midjourney in many ways, generating detailed, creative, and visually appealing imagery. It has shown significant improvements over the previous version of DALL-E.

  • What is the significance of the human brain processing visual information faster than text?

    -The human brain processes visual information about sixty thousand times faster than text, which is why Microsoft is focusing on creating visual tools. This is a critical aspect of how people search for information, and it enhances the user experience.

  • How can users access the new Bing image generation feature?

    -Users can access the new Bing image generation feature through Microsoft's new Bing chat feature or directly within Microsoft Edge, making it the first and only browser with an integrated AI-powered image generator.

  • What are some of the limitations or challenges faced with the new Bing image generation feature during the testing?

    -During testing, the user faced challenges such as the chat box not supporting images initially and having to wait for image generation, which could be slow without using a 'boost'. Additionally, the aspect ratio selection was not available in DALL-E 2, unlike in Midjourney.

  • How does the updated DALL-E 2 algorithm perform when generating complex prompts like an anthropomorphic lemon character?

    -The updated DALL-E 2 algorithm performed well with complex prompts, generating a coherent and detailed anthropomorphic lemon character. The results were creative, combining styles and ideas effectively, and were considered better than the previous DALL-E results.

  • What is the role of Microsoft Rewards points in the new Bing image generation feature?

    -Microsoft Rewards points can be redeemed for more 'boost' credits if users run out. These boosts speed up the image generation process. Users can earn these points by participating in quizzes or daily polls.

  • How does the new DALL-E 2 algorithm handle more complex and specific prompts like a 'Walter White Lego character'?

    -The new DALL-E 2 algorithm significantly improved from the original version, providing more coherent and detailed images that closely matched the prompt. However, it still did not reach the level of detail and realism achieved by Midjourney V5.

  • What are the differences between the new DALL-E 2 algorithm and Midjourney V5 in terms of image quality and realism?

    -While the new DALL-E 2 algorithm has improved greatly and produces high-quality images, Midjourney V5 often generates images that are more photorealistic, with clearer details and better coherency. Midjourney V5 is considered to have an edge in creating images that closely resemble professional photography.

  • What are the benefits of the new pricing model for the Bing image generation feature?

    -The new pricing model offers technically infinite image generations for free, with the potential for slower generation times if users do not use a 'boost'. This model is more favorable compared to the original DALL-E 2 pricing structure, providing more value to users.

  • How does the new DALL-E 2 algorithm perform with prompts that require combining different elements to create something new and fun?

    -The new DALL-E 2 algorithm shows significant improvement in handling complex and creative prompts, generating detailed and artistic images that combine different elements effectively. It demonstrates the ability to understand and visualize abstract concepts, although it may still have some limitations compared to Midjourney V5.

Outlines

00:00

🤖 Microsoft's Integration of Dolly Algorithm in Bing

The video discusses Microsoft's collaboration with OpenAI to incorporate the Dolly algorithm into Bing. The Dolly algorithm, previously featured in a video, is known for generating high-quality images. The host expresses excitement about trying out the new Dolly 2 algorithm, which is now accessible through Microsoft's Bing chat feature. The video also mentions that the Dolly algorithm has been used in over a hundred million chats with the Bing AI, which employs gpt4. Despite some initial issues with generating images through chat, the host appreciates Bing's ability to find relevant images. The video highlights the potential of the Dolly algorithm in creating visual content and the integration of Bing Image Creator into the browsing experience.

05:02

🖼️ Comparing Dolly 2 and Bing's Image Generation

The host compares the image generation capabilities of Dolly 2 and Bing's new image generation tool. After generating images of crocodiles and lemons, the host notes that Bing's Dolly 2 produces more detailed and realistic images than the original Dolly 2. The video also explores the use of 'boosts' in the image generation process, which can speed up image creation. The host concludes that while Bing's Dolly 2 has improved, it still falls short of the quality produced by Mid-Journey V5, another image generation tool.

10:02

🐱 Image Quality and Realism in Dolly and Mid-Journey

The video continues with a detailed comparison of image generation tools, focusing on the prompt of a black and white cat. The host finds that while the new Bing Dolly algorithm produces good results, Mid-Journey V5 still outperforms it in terms of image quality and realism. The host also notes that Dolly 2's generation times are faster, but the quality of images from the new Bing Dolly is notably better, with fewer blotchy and incoherent results.

15:04

🧩 Dolly 2's Improvements Over Original Dolly

The host examines the improvements in the new Dolly 2 algorithm over the original Dolly. Through a series of image generation prompts, including a 'Walter White Lego character' and a '1940s detective frog,' the host observes that the new Dolly 2 produces more coherent and detailed images. However, it still does not match the quality of Mid-Journey V5, which maintains a higher standard for image generation.

20:06

📸 Realism and Character in New Dolly 2 Generations

The video showcases the new Dolly 2 algorithm's ability to generate images with a high level of detail and character. The host generates a variety of images, including a 'professional photo of a shitzu,' a 'skateboarding penguin,' and a 'shitsu on a pirate ship,' noting that the new Dolly 2 provides more realistic and artistic results compared to the original version. The host also attempts to recreate a logo with the algorithm, resulting in a series of sharp 3D renders.

25:07

💬 Viewer Engagement and Conclusion

The host invites viewer opinions on the new Dolly 2 model and its comparison to Mid-Journey V5 and other models. The video concludes with the host's appreciation for the free and unlimited image generation capability of the new Dolly 2, despite the potential for longer generation times without 'fast credits.' The host also acknowledges the advantages of Mid-Journey's subscription plans and thanks the viewers for their support.

Mindmap

Keywords

💡Bing AI Art Generator

The Bing AI Art Generator is a feature developed by Microsoft that utilizes the Dolly algorithm to create images based on textual prompts. It is integrated into the Bing search engine and is showcased as a competitive tool against other image generation platforms like Midjourney. The video discusses the capabilities and output quality of this generator, comparing it with other services.

💡OpenAI

OpenAI is a research organization that develops AI technologies. In the context of the video, Microsoft collaborates closely with OpenAI, particularly in incorporating OpenAI's Dolly algorithm into Bing's features. OpenAI's contributions are central to the advancements in AI-powered image generation discussed in the video.

💡Dolly Algorithm

The Dolly algorithm is an AI model used for generating images from textual descriptions. It is mentioned several times in the video as the underlying technology that powers the Bing AI Art Generator. The video provides a comparison of the newer version of the Dolly algorithm with its previous iteration and other platforms.

💡Midjourney V5

Midjourney V5 is an advanced AI image generation platform that serves as a point of comparison in the video. It is noted for its high-quality image outputs and is used to evaluate the performance of the Bing AI Art Generator. The video host expresses excitement about the potential of Dolly to compete with Midjourney.

💡Image Generation

Image generation refers to the process of creating images from textual descriptions using AI algorithms. It is the main theme of the video, where the host explores different AI models' capabilities to generate images that are coherent, detailed, and artistic.

💡AI Powered Visual Stories

AI powered visual stories are a feature that combines written and visual content, generated by AI, to provide a more engaging and informative experience. The video discusses how Bing integrates this feature to enhance user interaction and information search.

💡GPT4

GPT4 is an advanced version of the Generative Pre-trained Transformer developed by OpenAI. It is mentioned in the video as the AI model that Bing chatbot AI uses, which is significant for understanding the chat-based interaction with the Bing AI Art Generator.

💡Photorealism

Photorealism in the context of the video refers to the quality of AI-generated images resembling real photographs. It is an important criterion for evaluating the output of the Bing AI Art Generator and comparing it with other platforms like Midjourney V5.

💡Microsoft Edge

Microsoft Edge is a web browser developed by Microsoft. The video highlights that the Bing AI Art Generator is integrated into Microsoft Edge, making it the first and only browser with an integrated AI-powered image generator, which is a notable feature for users.

💡Creative Mode

Creative mode is a setting within the Bing AI Art Generator that allows users to generate images with a focus on artistic and innovative outputs. The video host attempts to use this mode to generate images but encounters some limitations.

💡Boost Credits

Boost credits are a feature within the Bing AI Art Generator that allows users to speed up the image generation process. The video discusses how these credits can be used and replenished, affecting the user experience with the image generation tool.

Highlights

Microsoft collaborates with OpenAI to incorporate the DALL-E algorithm into Bing, offering a new image generation feature.

The new DALL-E 2 algorithm is highly competitive with Midjourney, generating detailed and creative imagery.

Access to Bing Creator image generator is available through Microsoft's new Bing chat feature.

Bing's image generation can be accessed through a separate website or integrated into Microsoft Edge.

Bing AI uses gpt4, which is fascinating for its capability in generating visual stories and knowledge cards.

The human brain processes visual information significantly faster than text, making visual tools critical for information search.

Bing's image generation is initially rolling out in creative mode and is only supported in English.

The new DALL-E algorithm generates high-resolution images that are more realistic and detailed compared to the older version.

Bing's image generation is slower than DALL-E 2 but offers higher quality images.

Midjourney V5 outperforms both Bing and DALL-E 2 in creating photorealistic images.

Bing's image generation can be enhanced using Microsoft rewards points to speed up the process.

The new DALL-E model in Bing is free and allows for infinite image generations, albeit potentially slower without a boost.

Midjourney V5 is praised for its ability to create detailed and coherent images that closely resemble professional photography.

The updated DALL-E model in Bing shows significant improvement over the original DALL-E 2, especially in character and concept generation.

Bing's image generation offers a creative and artistic approach, with results that are more like 3D renders.

While Bing's new DALL-E model is impressive, it still falls short in comparison to Midjourney V5 in terms of image quality and versatility.

The new DALL-E model in Bing does not currently support aspect ratio adjustments, which were available in the original DALL-E 2.

The free and unlimited generation aspect of Bing's new DALL-E model is a significant advantage over the previous model's pricing structure.