The Real Threat of AI: Over-Attachment To How You Do Your Craft

This isn't about AI taking jobs. This is about people being over-attached to how they work, and that attachment becoming their (and their company's) biggest vulnerability.

John-Daniel Trask said to me the other day that most people are stuck on how they do their job, not the why they do their job. I believe this is the single biggest unconscious mental barrier and threat to business competitiveness in the AI era.

This has nothing to do with technology quality, but all to do with how we're optimising for efficiency instead of growth, for protecting the status quo instead of creating opportunity.

Dave Howden points out that humans built the Panama Canal by hand. Took 50 years. Cost 20,000 to 30,000 lives. When machines came along, we built the Industrial Revolution. The same shift is applying to knowledge work today but most people aren't thinking about it from that perspective or the real inefficiency in humans continuing to perform a large number of knowledge worker tasks that we do today.

The rude awakening people are going through is that they're attached to how they work. We see this when new people are added to our teams, or work is offloaded to someone else, or even when we get promoted and have to let go of certain tasks ("But I like that part of my role"). Our instinct is resistance. We've tied ourselves to how we do work, not why.

How we work encompasses craft we've refined over 10s if not 100s of thousands of hours.

That expertise feels like our identity.

When AI uproots how we work (from "I typed the code" to "I orchestrate agents", from "I wrote every word" to "I ideate alongside AI"), it feels like our foundation is crumbling.

This shift is emotionally brutal. When our identity is wrapped up in being "the coder" or "the writer", having AI do those tasks feels like erasure, creating that feeling of grief.

You can absolutely choose to keep working the manual way. No one's going to stop someone from typing every line of code, crafting every word by hand, building every spreadsheet formula manually. But moving forward, that level of activity is becoming a hobby or learning exercise. When we're doing it for a client or employer, it's business.

Confusing the two is where the pain lies. Despite the perceived ethical and morale considerations, it becomes a mismatch when our desire for manual craft is imposed on a business that pays us for efficiency and outcomes:

Now, there absolutely are businesses where manual, labour-intensive, handcrafted work is the commercial model: artisan furniture, bespoke jewellery, hand-stitched leather goods, custom craftsmanship that people buy precisely because it isn't automated. In those businesses, the process IS the product. The manual work is commercially justified.

But a good majority of businesses at scale don't operate that way. They sell uniform outputs, consistent services, reliable delivery, predictable quality, and repeatable results. In those environments, insisting on manual, labour-heavy processes (and yes, I'm including knowledge work here) isn't sustainable. Market pressure will erase that approach regardless of how attached we are to it.

This is the Rubik's Cube rotation some folks need. To stop looking at the "how I execute" face. Start looking at the "why I create value" face. You're not "the person who writes SQL queries". You're "the person who ensures decisions are made on clean data". The title stays. The outcome stays. The tools change.

This doesn't make someone any less human, it just requires a total rethink of how we achieve outcomes for us, a business and society.

Why we work can be broken down into several areas. Why we work should start with: what am I doing personally? What is my why for doing this job in particular or completing this task?

This manifests in important ways:

There was a goal. That was the why. How I got there is almost irrelevant. Of course, some will argue that how you got there is everything. I disagree completely. The how will always change. It is a concept that will change continuously. It has throughout history.

The how we get there is using a different tool or technology, thinking differently, working with different people. We will still be applying the same human-level judgment to a problem or situation. Coming up with a solution, whether it be with people or with technology, is just marking the homework to achieve an objective.

This shift requires a new answer: who are you?

You're no longer defined by the tools you hold (the keyboard, the spreadsheet, the code editor). You're defined by the vision you possess. Everyone shifts from the "hands" of the organisation to the "head", just operating at different levels of complexity.

The new high-value work is in framing problems correctly, choosing the right tools, evaluating outputs critically, adding judgment and taste. That's orchestration, not execution.

Our brains are processing units, not hard drives. By clinging to manual work (memorising, collating, sorting), we're using our brains for low-level compute tasks rather than high-level creative tasks. AI frees us to think rather than just process.

In a world where content generation is becoming free, curation becomes the premium skill. The ability to look at three AI-generated options and say "That one fits our strategy" is the new hard work. As the how becomes cheaper, the why becomes more expensive.

Some say, "I can choose how I work." Of course you can choose the old way: create formulas manually, type every word, copy-paste between 15 tabs. That's valid for hobbies, or maybe when you're in learning mode and getting to grips with a task. But when it comes to market forces, that's inefficient and non-competitive.

You will be railroaded by the next person who has figured out how to do your tasks with automation and AI at higher value to the customer (either better quality for the same price, or same quality at a lower price).

I know that sounds harsh. But the alternative (letting people believe their methodology will remain viable) is crueller. The adage of "losing your job to someone else using AI" will hit like a freight train. People spending three hours typing meeting notes plus an email versus people working with AI, completing tasks in 10 minutes with better results.

If you're feeling defensive, that's the signal. You're still attached to the how.

What's more concerning is people who think they've adapted are still stuck. The reality I'm seeing is that people believe they understand where AI can be applied to their workflow. They're using ChatGPT and Copilot to rewrite emails, analyse reports, help mock up designs. They think they're doing AI adoption. They're not even scratching the surface.

They haven't started thinking about how they can use tools like Google AI Studio to completely automate a process with natural language and prompting that doesn't need sign-off from their IT team. Tools that can fundamentally reshape the way they work.

They're stuck in the how: "How do I move sales meeting notes from a transcription tool into a CRM to draft an email for me?". When they should be asking the why: "Why am I worrying about this process at all? How do I just get meetings booked in my calendar without touching any of these systems?". The difference between those two questions is the difference between efficiency AI and opportunity AI. One saves you 20 minutes. The other redesigns your entire workflow, delivers an outcome, and frees you to do work that people care about.

Now, not everyone has access to these tools yet (it's only a small credit card payment away), and not every organisation is ready for this shift. But that doesn't change the direction of travel.

Executive teams have responsibility to employees but also fiduciary responsibility to shareholders to deliver the highest quality service at the lowest cost. As above, what most get wrong is optimising for efficiency when they should optimise for opportunity.

Holly Knill GAICD nails this: "Efficiency AI is a shortcut to nothing. It might buy you a year of EBITDA improvement, but you can't cut your way into growth. No company ever grew significantly through cuts and constant cutting".

There's a cost of doing nothing and that cost is borne by the employee. If a competitor automates these processes before you do, your whole business goes down. But the trap is that if you're just "using AI to save Kate and Dexter eight hours a week so they can spend it on TikTok and Netflix", you've completely missed the point.

All these capabilities exist today. They should be mandated by executives to implement. But here's the challenge: the executives and the knowledge workers are both getting stuck because they're not looking deep enough. They're not taking the time to experiment, be curious, or invest in specialist capabilities to show them where the opportunities are.

This requires understanding your data better, training your people to identify opportunities, actually building out those opportunities, and building in governance layers to continually refine where opportunity assessments are presenting themselves. That's the real work. Not asking ChatGPT to make your email sound more professional.

The question executives should be asking: if you're saving someone 20 hours a week, what are they doing with that time? Are they making more money for the company? Are they opening new markets, creating new products, serving new customer segments? Because that's where productivity growth actually comes from.

Don't confuse productivity with efficiency. They're not the same thing. And the trap with AI is that we're focused on doing the same thing faster, rather than automating the thing to focus on new activities that drive commercial outcomes.

The market doesn't care about your hesitation. It rewards the organisation that delivers better outcomes at lower cost.

The paradox is focusing on why instead of how makes you more valuable, not less.

When you're outcome-oriented rather than process-oriented, you become infinitely more adaptable. When you're outcome-oriented rather than process-oriented, you become infinitely more adaptable.

You're not tied to a methodology that might become obsolete in 18 months. You're the person who can solve problems regardless of what tools emerge. That's the ultimate job security. Not expertise in a single process, but the ability to drive results in any environment.

The professionals who thrive will demonstrate tool fluidity. The ability to pick up new AI capabilities and immediately apply them to drive business outcomes. Tool attachment is a liability. Tool fluidity is the new expertise.

Your value isn't in how you execute anymore. It's in the judgment you bring, the problems you frame, the outcomes you drive. That judgment becomes exponentially more valuable precisely because execution is now automated.

Start with one task you do regularly. Ask: what's the outcome I'm trying to drive? Then ask: is the way I'm currently doing this the most effective path to that outcome, or am I just attached to the process?

Try doing that task with AI assistance. Use AI to handle information processing, first drafts, research aggregation. Then apply your human judgment to shape it toward the outcome that matters.

List five tasks you do today. For each one, write the real why behind it. Then ask how you'd rebuild it from scratch with AI and automation. Notice what resistance comes up. Ask yourself: is this resistance protecting something valuable, or protecting something comfortable?

The question isn't whether your job will exist. It's whether you'll evolve to stay relevant within it, because AI is coming to your job whether you like it or not (shout out Jason Paris for this phrasing). The how is already changing. Will you detach from it quickly enough to refocus on the why, before the market decides for you?

Rotate the Rubik's Cube.

The how is negotiable. The why is not. Figure out which one you've been optimising for, and adjust accordingly. The market will adjust with or without you (thanks U2).