Google announced that it has enhanced the “Ask Photos” function’s speed in returning search results after momentarily pausing the distribution of its unreliable AI-powered feature in Google Photos.

Users may use natural language queries to search through their collection of digital photos thanks to the AI technology, which was first unveiled at Google’s I/O developer conference last year. When Ask Photos receives input, it uses Google’s Gemini to leverage the AI’s comprehension of a photo’s content and other metadata.

Users, however, claimed that the AI feature was unreliable and frequently took a long time to react while it was “thinking.”

Earlier in June, Jamie Aspinall, the product manager for Google Photos, addressed these issues by writing on X that “Ask Photos isn’t where it needs to be, in terms of latency, quality, and ux.” He also mentioned that the rollout would be halted for a few weeks while Google worked to restore the “speed and recall of the original search.”

Google announced in a brief blog post on Thursday that it is integrating the finest features of Photos’ traditional search function into Ask Photos, especially for basic queries like “beach” or “dogs.” As with traditional search, this enables the search results to appear more rapidly.

Meanwhile, the AI will be working in the background to identify the most pertinent images and attempt to respond to increasingly complicated questions.

For example, if you look for a picture of a “white dog,” a number of preliminary search results show up right away. The findings of the AI’s study will be shown below, along with some introduction language that might include your dog’s name if you entered it and the date that the animal’s images first surfaced.