Google is introducing a feature similar to Bing's DALL-E 3 AI image generation. Users can now generate images from text prompts directly within Google Search.
Strict content filters are in place to prevent misuse. Google Imagen powers this tool, previously accessible to a limited audience. Initially, it's opt-in and available in English in the US.
Users can specify the type of image they want, choose from four results, and download them as png files.
Google assures that it will filter out potentially offensive content, blocking photorealistic faces and mentions of famous people.
Initially, access is limited to users over 18, and generated images will bear watermarks and AI-generated metadata.
However, concerns about safeguard effectiveness remain, as seen with Bing and Meta's AI tools.
Google acknowledges the tool's imperfections and offers a feedback mechanism.
It's unclear if this feature will become paid, unlike Bing's free access with limitations and paid boosts.
Other features in Google Search Generative Experience (SGE) include a conversational mode and customizable content length and tone.
These features are available to users who opt in through Google Search Labs in the US, and you can explore AI art tutorials for more insights.
