Google Lens is a powerful tool, and it always has been. However, the push towards an AI-focused ecosystem has taken some core Lens features and blatantly made them worse. Artificial intelligence is meant to take burdens off the user, but in many ways, Lens has become far too bloated, loaded with AI-centric features that make its overall experience more cluttered and more complicated.

Search in Google Lens has become interpretation-focused

More complex doesn’t mean better

Google Lens used to be a fantastic tool that did the heavy lifting for you. The Search function would essentially process an image and then convert it into a search query. This means that the context for the search was dealt with exclusively within Google’s search engine. This approach may seem a little primitive compared to how the current feature functions, but it ultimately created a more user-friendly end result.

Google has been upfront with how its new approach differs from the old way of doing things. They now describe Gemini as the brain behind everything. Now, Gemini takes a look at the image, understands what it is, and then decides the context of your search for you. It’s a shift from finding an object in an image to defining a whole scene, and the result is that you often have to chat with an AI bot in order to really get the result that you’re looking for.

The point of Lens is to find simple results. If I point my camera at some sneakers, I want to search for the sneakers, which usually means ending up with shopping results. With the new approach, Google thinks I need an interpretation of the sneakers, what they are, who they’re made by, etc., and my original purpose is buried. The whole magic of Google Lens was that you didn’t need to articulate your search; now you have to do exactly that.

Homework mode is gone, and trustworthiness has fallen

AI hallucinations make answers harder to trust

Reading sloppy handwriting using Google LensCredit: Jorge Aguilar / How-To Geek

A big part of the usefulness of Google Lens also centers around the trustworthiness of the response. A big area where this is highlighted is in the now-defunct Homework mode. You used to simply point your camera at a list of questions, it would search for the text, and it would generate answers based on search results, giving you links to sources that you felt like you could trust, which was a core part of what made it so useful. This also meant you could instantly fact-check the information yourself by tapping on a link.

With the focus shifting towards AI, the experience has changed. Google has completely removed the Homework mode and has instead integrated it into the search function. Now, the first thing that you’re met with is an interpretation of your question based on results that have been filtered through Gemini. Not only does this make things more cluttered and less focused on the answer, with you having to add context in some cases, but it also brings up the issue of trustworthiness.

Google’s AI overviews have been accused of being inaccurate pretty frequently. The term ‘AI hallucination’ has become pretty well-known across the internet. The AI overview also doesn’t consistently offer links to the sources of its information, and we’re left to trust a machine that has proven itself to be unreliable.

Looking for similar images can push you into a conversation

Don’t adjust the search box too much

Google Search new AI mode on a phone.Credit: Google

There have been community reports that adjusting the search box can, sometimes, push you into the AI mode instead of simply searching for similar images. A user reported that when they attempted to search for similar images to one they found online, and they adjusted the search box, they were pushed out of the simple function where you search for similar images and instead moved to AI Mode. The user was still able to return to searching for similar images, but being forced into a conversational search completely defeats the purpose of Google Lens.

The integration of AI in this case made the user experience worse. In a mission to be more complex, Google has taken a fairly simple feature and overcomplicated it. This links back to our first point, where we discussed how a context-based approach isn’t as useful in a lot of situations.

The other problem with this is that if you keep the search box tightly clipped around the part of the image you’re searching for, Google will sometimes analyze the entire context of the image anyway. This can cause the overview to spit out irrelevant information that has nothing to do with your search.

AI in Google Lens has made things more complicated for the user, not less

The whole point of artificial intelligence is that it is meant to take the burden off the user. While this has a lot of useful applications, Google has essentially applied it to a tool that was already quite effective, overcomplicated things, and made the experience worse for its users, at least in some areas.

This combines with the fact that everything is more cluttered, direct information is harder to find, and sources are hidden and left for you to interpret anyway. Constant updates also move everything around and change the functionality of features, leaving users confused, and sometimes even when a feature isn’t removed, users think that it has been because it has been buried in a mountain of menus and AI overviews.