Generative AI and Its Potential Two-Fold Impact

This article, published on October 19, 2023, was written by Jerome M. Wendt, CEO and principal data protection analyst, DCIG, LLC.

Many organizations may currently view generatie AI as a buzzword being heavily hyped. Some may even view AI as too abstract and/or perhaps even too evil for use in business settings. There is perhaps some truth in these viewpoints.

However, I have started to examine how generative AI might potentially impact on data protection and data storage technologies. It quickly became apparent to me that its greatest impact may be two-fold. It stands to better equip organizations to manage these two technologies and more quickly adopt new features in them.

Generative AI seems too abstract (and too evil!)
Probably by now everyone involved in the technology industry, and even many who are not, may have experimented with AI. This experimentation may have involved getting a free account with ChatGPT and asking it a few (or a lot of) questions. While fascinating, many still see this use case as entertainment but too abstract for any practical use.

Other individuals have gone even further and tried to break ChatGPT and/or expose its nefarious (i.e., evil) purposes. I recall having read a few articles in the past year how some individuals have done this.

They got ChatGPT or other AI tools to say things like humans are bad and should be exterminated. While that observation might exaggerate my point a bit, these individuals set out to demonstrate the inherently evil nature of AI.

This idea in and of itself that “AI is evil” is certainly nothing new. AI has been used as a plot line and concept in movies and TV shows for at least 30 years.

In my case, I am a big fan of the X-Files from the 1990s. Even then, episode seven in the first season referenced AI, which was notably released 30 years ago this month. Entitled Ghost in the Machine, the computer in that episode becomes or achieves artificial intelligence. The humans who manage it see this AI as a threat and attempt to shut it down. In response, the computer murders those who attempt to do so.

This combination of AI being seen as too abstract or, if used, becoming evil, can turn people off to AI. However, AI is like any other tool, one that must be properly understood, implemented, and managed. Once one takes the time to do so, then practical, beneficial use cases of AI begin to emerge.

Simple explanation of how generative AI works
Generative AI may seem like a black box when one first looks at it. While it certainly has black box components to it, it does not magically reach conclusions and generate output.

Rather, it operates like every computer program. It requires data, computing, storage, and networking to perform its analysis.

The big difference? Generative AI requires magnitudes more data, computing, storage, and networking to produce its results. Further, it can ingest and interpret structured and unstructured data in the many formats in which this data gets stored.

This represents the “black box” component of each generative AI offering. Each one uses different algorithms and processes to ingest, understand, and interpret the data. Further, each one likely has access to and stores different types of data.

These hard and soft factors combine to influence how well any generative AI program works. The more hardware resources (computing, networking, and storage) it has, the faster it runs and the more data it can process. Then the more data to which it has access, and the better the algorithms it possesses, the more meaningful its results.

Generative AI’s potential 2-fold impact on data protection and storage management
Working primarily in the data protection and storage spaces, generative AI has yet to have a meaningful impact in these areas. Admittedly, some providers already use generative AI to analyze technical data about their products. They then use generative AI (often called predictive analytics) to identify potential problems and resolve them before they become problems.

However, with generative AI rapidly maturing, organizations should stop viewing it as either too abstract or too evil. They should start viewing it as a new tool they can use to improve the management of their data protection and data storage in at least 2 ways.

First, use generative AI to analyze one’s current data protection and data storage environments
Data protection and storage products generate multiple alerts, logs, and reports. However, who has time to thoroughly read and understand all of them, much less query them as to what they mean?

Generative AI can do this without putting an organization’s IT environment or the data in it at risk. While we are still collectively in the early stages of generative AI adoption, this represents a logical place to introduce it. So often organizations fail to understand their IT environment. This leads to many wasted hours identifying the sources of current problems as well as implementing the wrong solutions to them.

Second, use generative AI to make better data protection and storage buying decisions
Knowing when to fix, upgrade, or buy data protection and data storage technologies currently feels more like a guessing game than a practical business and/or technical decision. This largely stems from a lack of understanding on what the current real problems are and what the next best technology choice is.

Generative AI can help on both these fronts. Per my first point, it can document the current environment pulling data from internal resources. However, it can also ingest technical documentation from private sources that vendors make available to organizations under NDA. They can use this data to better compare their current environment to new technologies and features that providers offer.

In this way, organizations may find an upgrade or patch to their current system resolves their issue. Alternatively, they may determine a new purchase represents their best option. Either way, organizations start making more decisions based on fact, not assumptions and hearsay.

Generative AI gets back to fact-based decision making
Organizations have way too much data – both internally and externally – at their disposal to fully digest and understand it all. Yet today more than ever they need a good handle on it to make the best business and technical decisions for their IT environment. Unfortunately, to date, this has been almost impossible to accomplish in a meaningful way.

Generative AI stands poised to change this. Employing it appropriately, organizations can feed information into it they trust to get recommendations out of it they can also trust.

Is it foolproof and mature yet? Absolutely not. But is generative AI too abstract and evil for organizations to start to use? Also, absolutely not. These use cases for generative AI in data protection and data storage management are admittedly not here yet today. However, expect them sooner rather than later. The business and technical arguments for using generative AI to manage data protection and storage are simply too compelling to ignore.