In recent years, an almost countless number of corporate sectors have felt the impact of artificial intelligence (AI) and machine learning (ML). Wall Street’s adoption of this technology has been no different — especially within digital asset management (DAM) — though some top experts suggest that growth, in this regard, is only just beginning.With this notion in mind, the immediate future appears poised to expand upon a variety of lingering trends and potential developments in AI and ML’s widespread implementation.
These are a few of the most pivotal trends in AI/ML asset management moving into 2018 and beyond.
Perhaps the biggest drawing point of AI/ML is its ability to streamline potentially tedious tasks in creative new ways. In DAM, these benefits have become more prominent as the technology grows increasingly sophisticated. Now, DAM professionals are shifting toward innovative approaches in implementing both AI and ML, exploring the possibilities of image recognition tagging technology, keywording, and search engine optimization (SEO) — among other concepts and services.
Relating assets via facial recognition
With regards to the previous section, image recognition has emerged as a major focal point within AI/ML-driven DAM — specifically in identifying the metadata of related assets. Though this specific technology has been the subject of a long evolution — encapsulating years of troubleshooting and fine tuning — industry experts are now starting to debate its readiness for widespread application. Some commentators feel that AI/ML, sophisticated as it now is, is not yet ready to fully vehicalize DAM-related facial recognition, citing the continued risk of inaccuracy and the setbacks that could snowball as a result. Other observations argue that the technology is already on the cusp of full clarity and efficiency. Both parties do technically agree on one fact, however: even if AI image recognition is not yet ready to introduce to metadata flows, it will be in the near future.
Terminology and potential challenges
Though DAM-grounded AI/ML is undeniably on the upswing, there is still a shortlist of challenges facing its final phases of implementation — the main one rooting itself in potentially tricky asset keyword terminology. These issues may present themselves in systems operating atop a client taxonomy, in which asset keywords are commonly compiled in a “master list.” In order to efficiently work in AI/ML as a complement to this metadata, these systems would likely require a fair amount of redesigned coding and other organization changes — a problem that is not insurmountable, but definitely significant.
The aforementioned matters represent just a few of the considerations currently being taken in DAM AI/ML, as the future appears to hold quite a lot for this vital sector of business functionality.