Users of Google's artificial intelligence Bard chatbot can now create images, adding a feature that competitor OpenAI's GPT Plus has had since October last year, through the DALL-E 3 text-to-pictures generator.
The new capability is through Google's Imagen 2 model which can produce photo realistic results.
Following reports of users bypassing prompt safety restrictions on Microsoft's Designer AI tool to generate nude images of pop star Taylor Swift, Google is at pains to state image generation with Bard "was designed with responsibility in mind".
Bard also adds the SynthID embedded digital watermark to images, to distinguish between machine generated and original human artwork.
Imagen 2 also powers the standalone ImageFX image generator, part of Google's experimental AI Test Kitchen that's available for users in New Zealand.
The AI Test Kitchen also offers the TextFX language tools, which Google says is made in collaboration with rapper Lupe Fiasco.
Another standalone AI tool, MusicFX, lets users create music by entering text prompts.
First available in September last year in English only, Google has now added the over 40 languages to the double-check responses feature in Bard.
Users of Bard can substantiate the chatbot's responses with double-check which searches for content on the web to evaluate and learn more about the information found, if any.
Earlier on, Google chief executive Sundar Pichai said the company's Bard Advanced, which will be underpinned by its most powerful Gemini Ultra model, will be a subscription service.
Pichai made the announcement in a Google earnings call, but did not say how much the subscription charge would be, or when exactly Bard Advanced would be available.
Competitor ChatGPT Plus is priced at US$20 a month currently.
Apart from Google itself, Samsung has signed up to use the Gemini generative AI.
Google's Gemini Nano is included in Samsung's latest flagship smartphone Galaxy S24, through the Android 14 mobile operating system. The Galaxy S24 also features Imagen 2 for image generation.
We welcome your comments below. If you are not already registered, please register to comment.
Remember we welcome robust, respectful and insightful debate. We don't welcome abusive or defamatory comments and will de-register those repeatedly making such comments. Our current comment policy is here.