Google Adds Image Generation to AI Search, Ushering in the Era of Large Multimodal Models
Google has added image generation to its generative AI search (SGE), which is currently being tested. This move is in response to Microsoft’s recent addition of ‘Dali 3’ to ‘Bing Chat’. Both Google and Microsoft are racing to develop AI-powered search engines that can generate images in addition to text, in order to meet the growing demand for multimodal search experiences.
Image generation is a powerful new capability for search engines. It allows users to search for images using natural language queries, and to generate new images based on their descriptions. This can be useful for a variety of tasks, such as finding inspiration for creative projects, researching new products, or simply learning more about the world around us.
Google’s new image generation capabilities are powered by Imagen, a large multimodal model that has been trained on a massive dataset of text and images. Imagen is able to generate realistic and creative images from text descriptions, and it can also be used to translate images from one style to another.
Microsoft’s Dali 3 is also a large multimodal model that is able to generate images. However, Dali 3 is still under development, and it is not yet as capable as Imagen.
The addition of image generation capabilities to search engines is a significant development. It marks the beginning of a new era of multimodal search, where users will be able to interact with search engines in more natural and intuitive ways.
What does this mean for the future of search?
The rise of large multimodal models like Imagen and Dali 3 is likely to have a major impact on the future of search. These models can generate realistic and creative images from text descriptions, which opens up new possibilities for multimodal search experiences.
For example, users will be able to search for images using natural language queries, such as “a photo of a cat sitting on a couch” or “a drawing of a futuristic city.” They will also be able to generate new images based on their descriptions, such as “a photo of a cat sitting on a couch in a futuristic city.”
This could revolutionize the way we search for and interact with information. For example, instead of having to type in a series of keywords to find an image, we could simply describe what we are looking for in natural language. This would make search more accessible to people with disabilities, and it would also make it easier for people to find the information they need quickly and easily.
The era of the LLM is gone
The rise of large multimodal models marks the end of the era of the large language model (LLM). LLMs are AI models that are trained on large datasets of text. They can be used to generate text, translate languages, write different kinds of creative content, and answer your questions in an informative way.
However, LLMs are limited in their ability to generate and understand multimodal content, such as images and videos. Large multimodal models, on the other hand, are able to generate and understand both text and multimodal content.
This makes large multimodal models more versatile and powerful than LLMs. As a result, we can expect to see large multimodal models being used in a wider range of applications, including search engines, social media platforms, and creative tools.
Conclusion
The addition of image generation capabilities to search engines is a significant development. It marks the beginning of a new era of multimodal search, where users will be able to interact with search engines in more natural and intuitive ways.
The rise of large multimodal models like Imagen and Dali 3 is likely to have a major impact on the future of search. These models can generate realistic and creative images from text descriptions, which opens up new possibilities for multimodal search experiences.
The era of the LLM is gone, and the era of the large multimodal model (LMM) is here.