Steam’s recent annual Summer Sale was a bit of a mess this year, one of the functions that failed was the recommendation feature, that failed to recommend new games to individual players. It seems as a result of this, Valve is rolling out a brand-new service to help correct this and it’s called Steam Labs.
Steam Labs is set to use AI tracking tools to scan playtime and which type of games are currently in their library in order to recommend what to buy. This test service has three active parts or experiments: micro trailers, an interactive recommender, and an automatic show. Here’s the official description of the service from Steam’s website, “Every year, we create dozens of experiments around discoverability, video, machine learning, and more. You know who we thought might enjoy seeing them? Everyone.”
Micro trailers will present six second trailers that are designed to quickly inform players on what each game is about. If the trailer looks good, you can click on it to visit its page within the store. The automated show is a daily 30-minute video, or less, where Steam shows off a variety of game clips within this time. Finally, there’s the interactive recommender, this uses playtime history and your current library to make recommendations on what you might be interested in playing next.