Google introduces advanced AI capabilities to advertisers through Gemini integration
[ad_1]
Google announced on Thursday its plans to enhance its advertising products with the integration of its Gemini artificial intelligence models, aiming to offer advanced AI capabilities to a broader range of its customers. According to a Reuters report, this move is part of Google’s broader strategy to incorporate generative AI technology, capable of producing text and realistic images, into more of its services. Following the announcement that Gemini would be applied to Google’s enterprise tools, the tech giant revealed that the AI model would also augment Performance Max, a product designed to optimize ad placements across Google’s platforms, including email, search, and YouTube.
In an effort to make advertising more efficient and creative, Google introduced AI tools within Performance Max last November, enabling the automatic generation of images and text for ads. With the integration of Gemini, advertisers will now have the ability to create more engaging ads with extended headlines of up to 90 characters, a significant increase from the previous limit of 30 characters.
Additionally, Google’s image generation tool will be enhanced with its AI model, Imagen 2, allowing brands to create images featuring people. This development opens up new possibilities for advertisers, such as a florist quickly generating images of someone arranging flowers. Google’s Vice President of Search and Commerce Global Ad Solutions, Brendon Kraham, highlighted the potential of these tools to produce content in mere seconds.
To address concerns regarding the misuse of generative AI for creating misleading content or deepfakes, Google has implemented safety measures. These include restrictions against generating images of known individuals, celebrities, public figures, or content depicting violence or illegal activities. Furthermore, AI-generated photos will be watermarked to clearly mark them as synthetic images, ensuring transparency and preventing potential misuse.
[ad_2]
Source link