Bonus: Preparing for The AI Future
IN THIS SECTION, YOU WILL: Understand that balancing curiosity, doubt, vision, and skepticism is essential for driving sustainable innovation and change in organizations.
KEY POINTS:
- Architects must evolve from diagram-makers to decision-makers.
- AI amplifies productivity but demands more human leadership.
- Agency, judgment, and persuasion will define architectural impact.
What if the next architecture diagram you design is created in seconds—by a machine smarter and faster than you? AI is changing the landscape for everyone. If you’re an IT architect, you’re not in an untouched zone. In fact, you’re standing at a crossroads.
Jakob Nielsen recently made a compelling case that, in the age of AI, the job skills that matter most are agency, judgment, and persuasion. He was speaking to UX professionals, but his message couldn’t be more relevant for architects.
Why? Because when AI can generate ten architecture diagrams faster than you can open Lucidchart, your value isn’t in producing diagrams anymore. It’s in what you do with them. It’s in guiding teams through ambiguity, trade-offs, and implementation in the real world.
Inspired by Jakob Nielsen’s article, let’s explore what this AI transition means for your role as an architect—and what to do before a large language model takes your chair.
Agency: The Era of Waiting is Over
There was a time when architects could wait for requirements to roll in, produce a reference model, and move on. That time has passed. AI doesn’t wait. Neither do modern product teams.
The architects who are likely to thrive wit AI are those who take initiative. They create momentum, not wait for it. They don’t need permission to lead change.
Example: A mid-size logistics company was seeing inefficiencies across its microservices—latency, duplication, poor observability. The architect didn’t wait for a formal initiative. She used AI tools to scan their system design, identified patterns of redundancy, and built a prototype using an event-driven model with out-of-the-box observability. She didn’t ask if it was her job—she made it her job. That’s agency.
To assess your agency, ask yourself:
- Are you sparking conversations about how AI will reshape your tech stack or delivery model?
- Have you tested AI-driven design tools to cut down repetitive effort in your own workflow?
- Are you advocating for architectural changes that align with new business needs—even if nobody asked yet?
Having agency means moving from “solution provider” to strategic instigator. AI will handle grunt work. You bring the leadership.
Judgment: When AI Gives You 10 Options, Can You Choose the Right One?
AI is prolific. It will hand you ten plausible designs, complete with pros, cons, and performance metrics. But which one works here, now, with your people? There is lots of sensitivity, history, irrationality in organizations. AI doesn’t know the nuance. You do.
Example: A retail platform team used AI to generate three valid architecture proposals for integrating a new loyalty engine. One used a pub-sub model with Kafka, another proposed embedding logic into their existing CMS layer, and a third used API gateway orchestration. On paper, all three could work. But the architect knew the operations team lacked Kafka experience and the CMS team was overloaded. He chose the API-based design, minimizing risk and change management friction.
This is judgment: technical analysis grounded in organizational reality.
To assess your judgment skills, ask yourself:
- Can you look at five viable AI-generated options and quickly eliminate the ones that won’t fly in your context?
- Are you connecting design choices with business constraints, team skills, and cultural readiness?
- Do you balance performance optimization with long-term maintainability—even if it’s less flashy?
Judgment turns outputs into outcomes. Without it, all you have is… more diagrams.
Persuasion: Influence Is the New Essential Skill
Being right isn’t enough. We’ve all seen smart, solid architecture proposals collapse in a meeting because nobody bought in.
In flatter, faster-moving organizations, architects don’t get to “hand off” decisions anymore. You need to persuade. Influence. Evangelize. And not just to engineering leads, but to legal, finance, operations, even marketing.
Example: An architect at a SaaS company proposed migrating key services to a modular platform model using containers and service meshes. Engineering was sold. Leadership wasn’t. So he reframed the proposal—not as a technical upgrade, but as a way to reduce time-to-market by 30%, improve SLAs for premium customers, and simplify compliance reporting. Suddenly, everyone was listening.
This is persuasion: making the right decision make sense to others.
So, to practice your persuasion skills, try this:
- Replace “Kubernetes is more scalable” with “This lets us launch new customer features two weeks faster.”
- Translate “reduces tech debt” into “frees up 30% of developer time for revenue-generating features.”
- Use visual storytelling or simulations to explain the impact—not just the architecture.
- Your diagrams are only as useful as your ability to sell the story behind them.
Tools change—your ability to influence does not.
What Should You Do Right Now?
Jakob Nielsen suggests we have five years—at most—before current skills start to resemble blacksmithing in a world of self-driving cars. So what can you do today to get ahead? Here are some ideas.
Continuously Automate Yourself First
Stop manually drawing cloud diagrams or documenting dependencies line by line. Use AI to generate templates, compare patterns, or write governance policies. Use tools like ChatGPT or GitHub Copilot to draft RFCs or produce reference architectures from a prompt. Partner with AI—it’s your new intern.
Develop Judgment Intentionally
Start collecting real examples of AI-generated designs and review them critically. What’s missing? What’s unscalable? Build decision trees. Run architecture “tabletop exercises” with peers to hone your instincts.
Practice Persuasion in Small Moments
Try explaining your architecture choices to someone in sales. Or your CFO. Can you make them care? Can you translate abstract tech into tangible business value? Don’t wait for the big presentation. Influence happens in hallway chats and Slack threads.
Stay Grounded—But Look Up
Keep your fundamentals sharp (cloud, security, scalability), but widen your lens. Read about AI governance, ethics, and team dynamics. The best architects of tomorrow will be technical enough to build trust and broad enough to guide business impact.
Bonus Lessons from the AI Transition Era
Jakob’s article goes beyond skills. It’s a roadmap for changing your whole mindset. Here are five more shifts that matter deeply for IT architects:
Don’t just automate—reinvent.
The goal isn’t to speed up old architecture processes. It’s to invent entirely new ways of working with AI. Think: self-evolving architectures, design-time AI co-pilots, or live prompt-driven platform builders.
Be a first mover.
Nobody knows what AI-native architecture looks like yet. That’s your chance. If you invent and test now, you’ll be ahead of those waiting for consensus.
Choose your environment wisely.
Working at an AI-native or AI-first organization will accelerate your learning. If you’re stuck in an “AI-denying” culture, your growth—and relevance—will stall.
Let go of dead skills.
Mastering legacy frameworks or clinging to outdated governance models won’t help. Be willing to trade technical prestige for future-proof relevance.
Prepare for “pancaking.”
Teams are flattening. The future architect is not a gatekeeper, but a guide. Your ability to influence across flat, autonomous teams will define your impact.
Final Thought: You’re Not Being Replaced—Unless You Refuse to Evolve
Nielsen closes with this thought: “Around 2030, your old skills will be like being an experienced mammoth hunter.”
That might sound dramatic—but think about it. AI will likely write cleaner code, compare architectures, even predict system bottlenecks. But it won’t know that your customer ops team is burned out, or that your sales lead wants a new feature by next quarter, or that your team is already quietly resisting the platform shift.
That’s where you come in. Your value is no longer in drawing boxes—it’s in drawing people together.
Architecture should always connect technology with people, outcomes, and change. AI is a powerful tool. But leadership is still a human job.
To Probe Further
- Use the AI Transition Period to Transition Your Career, by Jakob Nielsen, 2025
Questions to Consider
- What parts of your workflow could AI take over today?
- How are you preparing to guide AI-generated outputs with human context?
- When was the last time you had to convince a non-technical stakeholder to back an architecture choice?
- How would you reframe a technical decision in business terms?
On Human Complexity |
|||
| ← | → | ||