I’ve gotta say, this one touches a kinda painful nerve with me.

10 years ago I walked into the Google Photos team expecting normal humans to do things like say, “Show me the best pictures of my grandkids.” I immediately felt like a fool: something like 97% of daily users don’t search, preferring to simply launch the app and scroll scroll scroll forever.

A decade later, the Photos team is talking about using large language models to enable uses like the following:

With Ask Photos, you can ask for what you’re looking for in a natural way, like: “Show me the best photo from each national park I’ve visited.” Google Photos can show you what you need, saving you from all that scrolling.

For example, you can ask: “What themes have we had for Lena’s birthday parties?”. Ask Photos will understand details, like what decorations are in the background or on the birthday cake, to give you the answer.

Will anyone actually do this? It’s really hard for me to imagine, at least as it’s been framed above.

Now, what I can imagine working—in pretty great ways—is a real Assistant experience that suggests a bunch of useful tasks with which it can assist, such as gathering up photos to make birthday or holiday cards. (The latter task always falls to me every year, and I wish I could more confidently do it better.) Assistant could easily ask whose birthday it is & on what date, then scan one’s library and suggest a nice range of images as well as presentation options (cards, short animations, etc.). That kind of agent could be a joy to interact with.