Efficiency has long been the ultimate badge of honor for technologists. Early in my career, the surest way to invite scrutiny was by posting your code to a forum and allowing others to comment. Whatever you did in 15 lines, someone else would do in nine, yet another in seven and some other would manage it in two. It was efficiency that was the badge of honor, not codifying the function in the first place.
AI can be the next iteration of that. Taking the efficiencies we achieved back in decades past and supercharging them by 100x. But while AI brings with it immense value and real opportunities to streamline work and improve operations, many continue to squander its potential and recommit sins of the past.
What happens when the money runs out
Costs around AI continue to climb, as memory costs soar, agentic workflows grow more complex, and organizations continue to look for new ways to incorporate AI to improve workstreams, offset limited resources, or augment security.
It is estimated that in the private sector, one in every four dollars spent on AI currently goes to waste. I believe that’s being optimistic. Federal agencies facing tightening budget pressure certainly don’t have that kind of money to burn. Even still, hidden inefficiencies risk costing organizations even more as costs compound and auditing is nearly impossible. If you’re not taking a strategic and measured approach to AI adoption today, it will only get easier to find yourself well outside of your budget six months down the road with little to show for it.
Here’s another way to think about it:
Years ago, I had a farm with lots of maple trees. Dozens of them. Naturally, I decided to try my hand at making maple syrup. I tapped all my trees and researched how to turn sap into syrup. After collecting sap for a few weeks, I had close to 100 gallons on hand that needed to be boiled down. By the end of the endeavor (many hours spent stoking the fire, adding more firewood, and watching sap boil) the net result was this: a scant few jars of homemade syrup and I’d burned every stick of firewood I had on hand. Talk about low returns; I’d made the most expensive syrup in history.
That’s what too many organizations are doing right now with AI. Tapping more trees, consuming more energy, burning more firewood, and fueling AI with endless resources, all with little (or less than expected results) to show for their procurement, implementation costs, and hard work. The entire purpose of AI adoption is augmentation, efficiency and cost savings, none of which have been accomplished.
Finding the right balance
Here’s where organizations, particularly in the public sector, can more effectively tap into AI. First, make sure you have a clear understanding of where AI can actually drive operational improvements or make your teams more productive. Don’t just use AI for the sake of using AI. We talk about this a lot in IT; the best way to identify practical opportunities ripe for automation is to look for the rote, mundane tasks that eat up the bandwidth of your top technicians. Things like patch management, documentation, and ticketing, tasks that are volume-based and metrics focused, are all areas where automation can significantly help. Start with smaller use cases. Prove value and expand from there. Measurement is key and having clear metrics of what outcomes look like before and after AI implementation matters when it comes to justifying spend, assessing progress, and deepening investment. Without this, you’re only real accomplishment is burning firewood, so to speak.
Second, fail fast. Often, AI initiatives stall because it’s not the right technology, not the right use case, and not enough individuals on the team with enough knowledge, context, or confidence to call a project when you find yourself well off base. As AI costs climb (and climb they will as VC-backed subsidies fade) so too will the cost of failure. If you’re finding yourself veering off course, the best thing to do is pause and reassess. Otherwise, you risk throwing money at a problem that could soon become a bottomless pit.
Lastly, prioritize your people. We’re seeing the effects of this already in the private sector, a quiet rehiring of individuals who were let go in lieu of savings promised by automation. While AI can significantly enhance and augment human expertise, we’re nowhere near a place where it can serve as a full-scale replacement for people. The organizations and agencies that invest in education and enablement around AI, in addition to setting proper guardrails around safe and effective use, will be the ones that reap the most benefits of the advantages AI presents.
I believe that AI brings with it a tremendous opportunity, and the productivity and efficiency gains that technologists 30 years ago could only dream about. But to make AI investments scalable and long lasting (especially as pricing models change), it’s important for organizations to have a firm grasp of the areas that make the most sense to automate. In addition to the good judgement and human expertise needed to push back when you’re running out of firewood and producing syrup at a net cost of $500/oz.
Egon Rinderer is Senior Vice President of Global Enterprise and Public Sector at NinjaOne. With over 35 years of experience spanning the Department of Defense, Intelligence Community, and private sector, Egon has built a career tackling the most complex technical challenges facing modern enterprises. Prior to joining NinjaOne, he served as Chief Technology Officer at defense tech startup Shift5 and as Global Vice President of Technology and President of Federal Business at Tanium. Beyond NinjaOne, Egon advises the U.S. government on national security matters, supports multiple defense technology startups, and actively mentors the next generation of innovators through STEM initiatives in higher education.
Photo Credit: Kampus Production, Pexels
Leave a Reply
You must be logged in to post a comment.