For many creators, searching for the right stock asset for a project takes up valuable time. Even with search filters, finding the perfect video clip or music track too often leads to endless browsing. But with the advent of AI and machine learning technology, new opportunities to improve the search experience are more widely available than ever before. That’s why we’re excited to announce that Storyblocks’ on-site search is now enhanced by AI to help our members find the right assets faster.
How our AI powered search works
By incorporating OpenAI’s large language model into our on-site search engine, we’re able to show search results that go beyond your queries. Even if a term is misspelled or extremely specific, our AI is able to quickly understand what content is relevant and show more comprehensive results.
Along with using AI to help get you the best results from our library, we also use a proprietary algorithm that evaluates how new content is performing every four hours. On top of that, we have a learning model which leverages the BERT framework that quickly analyzes how members are engaging with our content. With these enhancements, our search results are always adapting to what our members like. That way, you’re always seeing fresh new content that meets your quality expectations and content needs.
Ways Storyblocks uses machine learning to improve your experience
Smart thumbnail selection
The thumbnails that you see for our video content need to be able to quickly show you what each clip contains. We’ve leveraged AI technology to help us identify which thumbnails are the most relevant so you can easily find the right piece of content. With many creators juggling multiple projects and working with tight deadlines, our smart thumbnails help you find the perfect clip for your projects faster than ever.
Prioritizing representation
Since launching Re:Stock, we’re always looking for ways to showcase more representation in our library. With our on-site search, an average of 24% of the videos shown on the first page of results feature Black, Indigenous, or People of Color (BIPOC) models.
Cross-asset pairing
In 2022, we launched our cross-asset recommendation feature. It uses member behavior data to suggest complementary assets like video, images, and music tracks to go with whatever asset you’re considering using. You can learn more about how we use machine learning to suggest music that pairs perfectly with video footage and more here.
Creating a curated, top-quality library that works for you
All of the work we do with AI and machine learning is to help our members find the perfect asset for their projects quickly. That way, you save valuable time and streamline your creative workflows. Start exploring our award-winning library today and experience the Storyblocks difference for yourself!
