Ever since Gmail shook up the free online storage game in 2004 by giving away a then-shocking gigabyte of space, individuals and businesses alike have become "digital hoarders" -- keeping all their old data around in case it becomes useful later -- regardless of the unacknowledged drawbacks of the practice. And while storage has become exponentially cheaper over the last 15 years, the hidden costs of digital hoarding may finally outweigh their benefits.Google gave away a gigabyte in 2004 (and gives away 15 gigabytes today) because it wants Gmail users to archive every message, rather than deleting old emails and attachments. This provides both Gmail users and Google with unending access to all historical Gmail data, which the user can search through to find old communications, and which Google can use to train its machine-learning algorithms to better deliver search, spam filtration, and ad-targeting.
Unfortunately, for those of us that have 15 years of Gmail data, simply searching through our mail archives can be a tedious, frustrating exercise as we struggle to recall exactly who sent us what data when, so we can conjure up a search query that returns useful results. Our years of Gmail data is worth more to Google than it is to us, many days.
This same situation is playing out across the enterprise, as companies hurl enterprise search tools at their mass of disparate and disorganized data and documentation tied up in their various local hard drives and online file shares. Years or even decades of data from hundreds or even thousands of users makes for a lot of documentation to sift through in the hopes of delivering value. Simply indexing everything doesn't make it useful, as users often don't know what to search for or how to structure a query to find it.
Many organizations want to retain their data for the same reasons as Google -- to train artificial intelligence. What many companies are finding out is that this disorganized data is no more useful to AI software than it is to your employees. AI needs refined, annotated data to learn from, not information randomly strewn across many locations and systems with no context, organization, or version history. Without careful curation, AI can't make heads or tails of your data archives.
It's not enough to store everything, you have to store everything the right way. That's where next-generation knowledge bases come in.
An advanced knowledge base doesn't just store all your historical data in a single taxonomy. Modern knowledge bases guide you to annotate both old and new documentation to make this data both search- and AI-friendly, so that your information can actually be found, understood, and employed for maximum benefit.
Whether you're moving your old data into a self-organizing and annotating knowledge base or creating new documentation with annotation built in, a modern knowledge base ensures that your super-cheap digital storage doesn't end up costing you in the long run.