Yes, Be Afraid of AI, But Not for the Reasons You Think

AI is going to change your life. This isn’t an article about human’s aversion to change. This is about how we fail to learn from our mistakes and subsequently make our lives far more difficult than they should be, even in (or especially in) those changes that are meant to help.

I recently read a study summary report shared by Dave Wright of ServiceNow that looked at the early adopters of AI and how their behaviors matched up against the rest of the field, specifically on the NOW Platform (find the report here: https://lnkd.in/dkp2tpEP ). What struck me about that analysis while at the same time not surprising me at all? That neither group - neither the pacesetters nor the core - showed even basic proficiency in governance as demonstrated on the radar map (aka spider web chart).

I think early adoption of new tech is outstanding. I even do it myself sometimes (though, I do admit I still refer to recording things as “taping” them). It can be scary and there aren’t a bunch of others from whom you can learn or ask for help. But there is history from which to learn, even if our brains don’t do that math to see what even semi-related learning moments we can apply to this new tech.

Some would argue that we’re only barely on the downslope of the Cloud adoption curve, so it’s fresh in our minds. One of the key issues with Cloud Computing was what was known as Cloud Sprawl. Early adopters (aHEM) dove head first into the Cloud, using their company credit cards to spin up computing resources, leveraging resources to show how elastic computing resources could be with nominal investment (and none in hardware). These people were lauded for their contributions, whether they were quicker time to value for new solutions or the ability to flex for business seasonality or whatever. It was great, and a great resource to use.

Except, then, organizations needed to put some structure around all the fun that had happened on the cloud while they were focused on their good ole technology solutions. How much were they spending in the Cloud? What resources were being used for what services or applications? Who was paying for what and how reliant were key business services on someone’s credit card? What, exactly, is being done in the Cloud vs. on-prem? Do we know if we could move a whole application to the Cloud and what that would look like? If there’s a analogy closer to “herding cats” for the effort at reeling this sprawl back in, I’m might be “the horse is out of the barn,” but with hundreds of horses and a handful of sheep thrown in for good measure.

Organizations that have controlled Cloud Sprawl have done so by laying in a governance model for new things going into the Cloud.

They’ve defined architectural standards, maybe picking what Public Cloud to use in what circumstances or designating what can be done in the Cloud and what can’t. They’ve probably laid in standards around tagging and enforcing data protection. And a girl can always hope they have a way to keep this understood in the CMDB, specifically with service-awareness, to aid in incident troubleshooting and resolution as well as managing changes (as is required by any compliance framework you can think of that touches IT).

They have also defined a model to handle demand, where groups want to consume Cloud resources or challenge architectural standards, looking for an exception, or even laid out ways to assess pros and cons for doing something on-prem vs. in the Cloud. Any number of ways to manage Cloud demand in a structured way.

And these layers of governance are guided by a Steering Committee, responsible for overall strategy and alignment to and delivery on goals and objectives for the organization. They’ll promote what alignment to a ‘’Cloud First” strategy means after some C-level executive brought that phrase back from some conference they attended (iykyk).

Does this sound familiar to anyone outside of its application to Cloud Sprawl? Maybe because for as long as I’ve been in the ServiceNow ecosystem, a number of us have known a clear and decisive success factor in owning, managing, and driving value out of a wonderful, complex - but simplifying if you use it right - platform like ServiceNow is having a good, mature governance model.

Good governance actually speeds up time to value. I know, I know, the garage coders out there are all raising the wagging finger at me. I didn’t say, “Time to deploy stuff you can pass off to Operations and make it their problem.” I said time to value, and I meant not only for the organization as a whole, but for Operations. Things are now done in a structured way with high reliability, can be well supported, will be easily built off of and expanded, and as importantly, will be retired in a structured, smart way. That there is the value everyone hopes IT will deliver.

A million years ago, when ISACA first introduced the Certified in the Governance of Enterprise IT (CGEIT) certification (don’t look at me like that - it’s their naming I’m trying to be precise with), I jumped at it, attaining my CGEIT in their very first testing class (we can talk about why I didn’t maintain it over a beverage of choice). I have a love and passion for governance. It’s in my DNA, I think, as I’ve always been about method and process and decision-making and alignment to goals and, and, and. Sorry, I’ll stop. But my love of governance also comes from the same starting point as my love for technology. And for a good reason.

Without governance, you get chaos. Some chaos is more manageable than others, but that unmanageable chaos is what turns people off of technology. It makes it feel harder for non-technical folks to understand or embrace it. It makes them feel like this is tech only for the “big brains.” Over and over and over again, through history, technology without governance has created chaos (insert Jurassic Park quote here).

And without governance, those comfortable with the chaos and/or the technology will run as far as their legs will carry them. They’ll try all the cool new things and in all the cool new ways, plus a few we haven’t thought of yet.

In the AI space, what does this look like? I see a smattering of AI Notetakers in just about any meeting I’m in these days. One could argue they might provide a set of different summaries, some better than others at some things. But this assumes they all use different LLMs. If they’re all pointing at the same LLM or a couple of LLMs, then the only difference in the summaries is the font used. But one company has five to 10 different vendors being paid, losing out not only economies of scale but also the ability to ease non-techies into using an outstanding tool.

One also has to start wondering about these AI companies and how well they protect, manage, anonymize, and for that matter, understand your company’s data. Has anyone vetted these vendors? They have cool sounding names. They presented you with an 852-page EULA to click accept, so it has to be real, right? I mean, no one would deliver what could be proprietary information to Bob’s House of Data Playgrounds, would you? And is that vendor even conversant in what your business does or the outcomes you’re trying to drive.

When companies come to their senses, as they’ve done with Cloud Sprawl, BYOD, remote working, and so many other tech trends in the past, they’ll be charged with herding those cats. Even today, the companies best managing Cloud Sprawl might not be able to say authoritatively that they’ve accounted for everything ever done in the cloud before governance was put in place. This will be the same for AI.

And in the meanwhile, talking about AI or ServiceNow, how many opportunities have been missed? How many times has someone selected a tool, AI or a point-solution, because no one could help them easily adopt the AI tools already in use or the ServiceNow platform capability sitting right *there*?

A lack of governance is why you should be afraid of AI, not because it’s going to put us all out of jobs, become sentient, and rid the Earth of humanity (I mean, I hope it won’t do those things, either). And a lack of governance overall, in my opinion, is what holds companies back from success with cool tech like AI and ServiceNow. I have these governance conversations at least once per week in my job as a ServiceNow partner. I see posts in the ServiceNow subreddit that I have to sit on my hands to keep from typing, “Tell me you have no governance without telling me you have not governance for your ServiceNow platform.” Those who are successful have governance of some level in place, and those who wander aimlessly or founder don’t.

Do whatever you need to to lay in a good, thorough governance program in IT, overall and for big chunks of technology like AI (or Emerging Technologies, so you can handle whatever comes next) and ServiceNow. Use resources like the NOW Customer Success Center and ISACA or TOGAF or other governance models. Reach out to experts - partners, consultants, ubergeeks (I resemble that remark). Don’t ignore it - in fact embrace it - so you can drive value out of IT, investments, and your data the way everyone hopes you will.

AI isn’t coming. It’s here, and its call is coming from inside the house if you’re not careful.