AI is going to hack Jira
Hide your kids, hide your wife: how the Agile Industrial Complex has set us all up to be duped by AI codegen
In Big Agile, engineering = new features.
Whether you’re looking to track story points or adherence to roadmap or deployment velocity, there’s a “developer productivity dashboard” for you. In fact, there are about $200b worth to choose from, all promising to make you better, faster, and cheaper.
There’s just one problem: they don’t work.
Engineering productivity can’t be measured by tracking new features.
Engineering is not a new mobile app. Engineering is not deployments per day. Engineering is not a t-shirt size.
Engineering is building, maintaining, and evolving an interconnected system. This work of managing dependencies, resourcing, and architecture keeps you alive.
But in Big Agile, this existential work is invisible and taken for granted. Instead, you’re tracking byproducts.
Trying to manage engineering by looking at byproducts is like managing a company by only looking at margins.
Are you unprofitable because you’re investing in growth? Are you very profitable because you made reckless cuts? Absent context, it’s impossible to tell what a “good” margin is at all - are you a restaurant during the pandemic or a SaaS company in 2015?
As ludicrous as it sounds, this is the reality of how most companies manage engineering today. Tiny startups copy the DevOps of huge enterprises. PMs mindlessly allocate 20% of the backlog to maintenance - regardless of how much maintenance actually needs to be done.
This is all because the Agile Industrial Complex has seriously messed up the way we understand engineering, which in turn has transformed engineering management into a madhouse.
And AI is about to make it all much worse.
Because AI code gen absolutely excels at producing byproducts.
You can watch any number of flashy demos that show agents conjuring slick interfaces out of thin air. CEOs, founders, and investors are quick to push the narrative of human replacement. Every other LinkedIn influencer is trying to become a “one-man unicorn” - a billion-dollar company of one person, supercharged by AI.
I’ve written before on why the vibe is all wrong, but the tl;dr here is that AI is nowhere near true human replacement. The things an AI can do, while amazing for a machine, are trivial in the grand scheme of a career.
Ex: Building surface-level features is the equivalent of installing a sink or toilet in a house. A machine that can do this is certainly helpful, but it doesn’t encompass the field of plumbing or construction. You still need someone who knows what they’re doing in order to operate the machine at all.
This should be intuitive, but the Agile Industrial Complex has scrambled the very concept of engineering work.
Imagine if you believed that the work of building a house were a series of disconnected tasks - ex: wallpapering, installing toilets, and putting in new cabinets. Now imagine an AI machine that could do all of this for free, instantly.
Best of all, you don’t need to go through your very human, very expensive, (and often grumpy) contractor. You don’t need to negotiate on timelines, pricing, and specs. You don’t need to hear some tirade on “structural integrity” when you just want to try a new paint color.
Instead, you get to work with an ever-obliging AI agent who delivers endless permutations of anything you could ask for. Frictionless, free, and seemingly miraculous. You’ve long forgotten a simple truth:
Decorating is only possible if you have a standing structure to work with.
So you fire your expensive, grumpy human team and request larger and larger additions from your AI: a new guestroom, built-ins, and a walk-in closet.
You feel great, until…you realize that your new powder room doesn’t have running water; it was never connected to the water main.
You ask the AI to fix it. In doing so, it breaks the plumbing in the kitchen. It can’t tell you why because it doesn’t know. Neural systems are inherently black boxes; they can’t recognize their own hallucinations and gaps in logic.
There’s no “undo” button. While your AI was building haphazardly, your city made updates to the electric grid, water main, and fire safety regulations.
You reach out to your old human team…but they can’t help either. The house has changed too far beyond their recognition; they don’t know how things work anymore. Whether you choose to rebuild from scratch or troubleshoot literal hallucinations, you’re stuck.
This topsy-turvy tale is not a hypothetical, it’s already playing out in engineering orgs today.
Imagine a CEO who fires their highly-competent engineering team. Perversely, the high skill of that engineering team ensures that the organization will be just fine in the near term.
By definition, well-built foundations take a lot of work to screw up.
So nothing bad seems to happen, even though critical infrastructure has been cut. In fact, things seem amazing - EBITDA and “productivity” look great. No one notices the cracks forming in the foundation, especially now that its careful custodians are gone.
Instead, everyone prematurely toasts the CEO for a resounding success.
By the time the bill comes due, we’ve long lost the thread of cause/effect. Our CEO is long gone - off replicating his “success” at a new company.
This kind of leader has always thrived in environments like private equity, where margins are manipulated for short-term paper gains. But in this coming decade, they’ll find fertile pastures everywhere. We’re in an era of streamlining; everyone wants to cut bloat and “right-size”.
As a society, we’re all grappling with how to adapt to this new normal as worker bees, but engineering poses a unique, far-reaching threat:
The infrastructure of everything.
What will happen when we blindly apply Big Agile + premature AI tooling to core health services and patient data? Financial services? Social media and the spread of misinformation? Government infrastructure? Military infrastructure?
It’s 2025 and software engineering is the backbone of pretty much everything. But for all its prevalence and influence, most people don’t have the technical literacy to even grasp the basics.
It’s the most dangerous kind hype. We’re rushing to replace roles we don’t understand…with breakthrough technology we understand even less.
This may sound bleak, but the antidote is simple. It’s the same guardrail that virtually every other function outside of engineering has:
Common sense
Imagine if someone sold you an AI robot cleaner. You have an intuitive grasp of what the work of cleaning entails, which in turns helps you objectively evaluate how good of a replacement it actually is. When the AI starts putting paper plates in the dishwasher, you know it’s wrong - even if your favorite LinkedIn influencer says otherwise.
This common sense is what the Agile Industrial Complex has robbed us of.
Most non-technical leaders have never really engaged with the real work of software and systems management. They don’t know what it’s like to update a major dependency, complete a refactor, or learn a new language.
Instead, they deal in disconnected artifacts and ideas. The deeper you go into Big Agile, the worse it gets. Soon, before you can talk to an engineer, you’re “t-shirt sizing” your “user stories” before “poker planning” your “sprint” with your “scrum master”.
It’s the very definition of self-reinforcing bureaucracy: how else do you sustain a $200B industry that’s built around understanding and measuring engineering as a cost center and feature factory?
In different times, I likely wouldn’t so breezily recommend railing against the Agile Industrial Machine. But with the disruptive potential of AI, it’s not really a gritty rebellion at all.
It’s more like learning to surf a forming, strengthening wave.
Leveraging AI well in any field requires a practical, intuitive grasp of that field, and software engineering is no exception. Leaders who aren’t able to build this understanding will be played for chumps by the very methods they cling to.
Even if they don’t sabotage themselves off a cliff first, they’ll find themselves quickly drowned out by companies who are leveraging this breakthrough technology with finesse.
It’s just evolution, but still a tragedy. All that really set them apart was simple common sense.
Ready to surf/rage against the Agile Industrial Complex?
Christine Miao is the creator of technical accounting–the practice of tracking engineering maintenance, resourcing, and architecture.
Built on interviews with 200+ CTOs, technical accounting tracks the actual work of engineering rather than its byproducts. In just 15 minutes, everyone and anyone can build a common-sense, intuitive understanding of their engineering org.
Common sense isn’t that common.
And here’s the kicker: most people don’t care how something was made, only whether it works.
On the surface, a lot of things do “work, even when they’re held together with duct tape and wishful thinking.
As engineers, we care about what’s under the surface: the process, the pipelines, the architecture, the tests. But to most people, only the result matters, not the craftsmanship, not the automation, not the long nights debugging CI.
Kent Beck, one of the creators of Agile, described AI as an Unpredictable Genie immensely powerful, but prone to mischief if left unchecked.
And that’s our role:
To tame the Genie.
- Enforce checklists
- Write meaningful tests
- Be the quality gate.
You might not be the most popular person in the room.
But responsibility isn’t about popularity, it’s about doing the right thing, even when it’s invisible.
I enjoyed this post. This is the reason why I’m designing my own apps to the detailed level and only then letting AI write the code to match my specifications. I need to understand the design that thoroughly because it’s my duty to ensure that it’s solid. Ultimately, many are using AI as a means to abdicate responsibility for their work/roles and that isn’t sustainable for very long.