With the latest update to iOS (18.4), you’ll no longer need to read countless reviews before deciding whether to download an app. Instead, you’ll be able to just read the AI-generated review summary for that app before taking the plunge.
Apple Starts Summarizing Reviews Using AI
Anyone who regularly buys anything from Amazon (and that’s surely everyone reading this) will be aware of AI-generated review summaries. They sit atop the pile of reviews offering a summary, saving you from having to read the reviews in full.
How Amazon’s AI Review Summaries Will Make Shopping Decisions Easier
Trawling through reviews to verify a product is the right option can be tedious. But Amazon’s AI review summaries are here to make things easier.
And now, with the release of iOS 18.4 and iPadOS 18.4, Apple is following in Amazon’s footsteps. App Store review summaries are one of a number of new features outlined onthe Apple Developer website(as first spotted byMacworld).
As you can see from Apple’s example below, the review summaries will attempt to sum up people’s feelings about an app in a single paragraph using natural language. They should also highlight what people wrote in both positive and negative reviews.
Apple’s App Store review summaries are being generated using large language models (LLMs). The summaries will be refreshed at least once a week for apps that have garnered enough reviews to justify providing a summary.
Apple’s review summaries are currently only available to beta testers in English in the United States on a limited number of apps and games. However, all being well, they will become available to more people when iOS 18.4 rolls out generally in April. And then expand even further through the course of 2025.
Probably the most common use for AI right now is feeding an LLM a lot of information, and then having it summarize that information. So, with that in mind, it’s no surprise to see Apple using AI in this manner. Essentially, it’s a low-hanging fruit in the AI world; easy to engineer and with obvious, real-world benefits to users.
What Are Large Language Models (LLMs) and How Do They Work?
Generative AI is all the rage, but how does a large language model work?
AI Review Summaries Seem Easy to Game
The biggest issue with AI-generated review summaries is the ease with which they can be gamed. To be fair, all reviews can be gamed to a certain extent, with some unscrupulous app developers employing bots or paid shills to praise their products.
However, given that these AI-generated review summaries will pick up on the most common remarks, the potential for further gaming of the system is clear. Especially if they discourage people from reading detailed reviews (which may include more caveats and negative comments) in full.
