Google pauses 'Ask Photos' AI Feature to address performance issues

In a move that has raised eyebrows in the tech community, Google has decided to pause the rollout of its highly anticipated AI-powered "Ask Photos" feature within Google Photos.

Initially launched as an experimental feature last fall, "Ask Photos" leverages Google's Gemini AI models to offer users an innovative way to search through their photo libraries by asking natural language questions, The Verge reported.

The feature, which had been slowly rolling out to a select group of users, aims to allow individuals to query their photo collection for specific themes, events, or text-based elements.

However, product manager Jamie Aspinall revealed that the feature has been temporarily pulled back after feedback citing issues with latency, quality, and overall user experience.

"Ask Photos isn't where it needs to be," wrote Aspinall on his X handle, following an influx of criticism from users.

The product manager confirmed that the feature's rollout would be paused for "very small numbers" of users while the team addresses the concerns.

Aspinall emphasised that in about two weeks, a more refined version of the feature would be made available, one that would restore the speed and accuracy that the initial rollout had promised.

The pause follows growing scrutiny of AI-powered features, with Google's "Ask Photos" being no exception.

Powered by the latest version of Gemini AI, a specialised model designed specifically for this feature, "Ask Photos" is expected to understand and interpret the contents of photos, including text within images, to generate contextually relevant results.

But, as Aspinall noted, performance improvements are necessary to ensure a seamless user experience.

Alongside the pause of "Ask Photos," Google also made an announcement aimed at improving the overall search functionality within Google Photos.

From now on, users can perform more accurate searches by using quotes to locate exact text matches within filenames, camera models, captions, and even text embedded in photos.
Additionally, search results will now also include visual matches when quotes are not used.

This update, which expands on features announced earlier this year at Google I/O 2024, represents an evolution in the way Google Photos interacts with its users.

By incorporating multimodal AI capabilities, Google aims to make photo searches more intuitive and user-friendly, providing results that include both textual and visual context.

Google's decision to pause "Ask Photos" comes amid a rapidly intensifying competition in the AI space, where tech giants like Google, Microsoft, and startups alike are racing to develop groundbreaking new features.

Google's move to pause its AI-powered features is not unprecedented, as the company has previously taken similar actions when it identified flaws in its new technology.

For instance, after launching the "AI Overview" feature in Google Search last May, Google was forced to halt its rollout following multiple instances of nonsensical and inaccurate answers being shared on social media.

Earlier this year, Google's Gemini-powered image-generation tool also faced a setback.

According to The Verge, this happened after a series of reports highlighting historical inaccuracies in the generated images, such as portraying US Founding Fathers as people of colour, the feature was similarly paused to address these issues.

Despite the pause, the core vision behind "Ask Photos" remains intact: to leverage AI's growing capabilities to enhance the way users interact with their photos and videos.

By enabling users to ask natural language questions, the feature sought to transform the way photo libraries are organised and searched, offering a more human-like interaction than traditional keyword searches.

Google has yet to announce a precise timeline for the return of "Ask Photos" or additional details about how the updated version will improve upon its current shortcomings.

However, the company remains committed to refining the feature based on user feedback and further advancements in AI technology.