Google AI Gemini to receive real-time responses in app

Gemini, Google’s answer to ChatGPT, is set to get an update to its app that enables real-time responses, bringing it in line with how the artificial intelligence (AI) tool operates on desktops.

Discovered by AssembleDebug, an expert in Google apps, and reported on PiunikaWeb, the feature was found within the app’s settings.

✨ Gemini assistant to bag a new “Real-time responses” option (also a toggle to disable location access specifically for gemini)

📝 https://t.co/nv4NJYJ46K #Google #Gemini #AI #Android pic.twitter.com/A26BpMVvch

— AssembleDebug (@AssembleDebug) April 22, 2024

The real-time responses setting will mean that users can see responses to their prompts as they are being generated, rather than waiting for the response to be fully formed before being able to read it. This will speed up using the app, especially if a particularly long response is being created – not having to wait for the entire response to generate before being able to read it will be very useful for some users.

What is Gemini AI?

Gemini AI is Google’s generative AI suite. It is designed as multimodal from the ground up, meaning it can accept more than just written text as input – audio, images, videos, and code are all accepted by Gemini without having to switch between input types.

There are currently three types of Gemini. Gemini Ultra is the fully-featured flagship model; Gemini Pro is a lighter-weight version; and Gemini Nano is a smaller model still that is used on mobile devices.

Not only can Gemini accept inputs beyond text, but it was also trained on a variety of mediums, so theoretically it should be able to provide better responses than models that aren’t given such diverse training data.

Unfortunately despite bold claims and an ambitious strategy, Google Gemini has not been smooth sailing. The image generation features are currently paused due to inaccuracies in generated images. It has also come under fire for equivocating over responses to moral queries that some see as having definitive answers.

Regardless, Google continues to resolve issues and improve the model to ensure it does not lose too much ground in the generative AI race where it is head to head with OpenAI, Microsoft, and others.

Featured image credit: Google