Personally, I still invest in humans. Not in the sentimental corporate sense where organisations speak endlessly about valuing people while simultaneously stripping away institutional knowledge, flattening specialist roles, and treating experience as an inconvenient cost centre. I mean I genuinely invest my trust in experienced human beings — the ones who know how to think through complexity, recognise patterns, slow things down when necessary, and identify when a problem has been framed incorrectly from the very beginning.
That conviction has only deepened while watching the current AI frenzy unfold. Everywhere online now, there seems to be another advertisement warning professionals that they are moments away from irrelevance unless they immediately reinvent themselves as AI-powered productivity machines. The tone of many of these campaigns is not educational so much as psychologically coercive. They do not simply market tools; they market anxiety. The target audience is obvious: professionals in their forties and fifties who already sense the subtle shift in how experience is being treated within recruitment spaces.
Much of this advertising carries an unmistakable undertone of ageism. The videos are slick, fast-moving, and emotionally manipulative by design. A younger person types a few prompts into an interface while triumphant music swells in the background. Entire careers are supposedly compressed into a few automated clicks. Designers are replaced. Writers are replaced. Coders are replaced. Consultants, trainers, analysts, strategists, and project managers are all apparently standing on the brink of obsolescence because somebody discovered how to generate output quickly.
The implication is rarely subtle. Adapt immediately or become irrelevant. Learn prompting or prepare to be discarded. The message is not simply “here is a useful new tool”. It is “your accumulated experience may no longer matter”. That is not innovation literacy. That is fear dressed up as professional development.
You can feel the same logic creeping into interviews now. After decades of work, implementation, troubleshooting, governance, delivery, stakeholder management, political navigation, and operational accountability, experienced specialists are still expected to sit through oddly performative questioning rituals that seem entirely disconnected from the realities of senior work. There is something faintly absurd about a person with twenty or thirty years of complex systems experience being asked where they see themselves in five years, or how they plan to “stay relevant” in the age of AI, particularly by panels who themselves may never have stabilised a failing implementation under real-world conditions.
What makes it more irritating is the increasing tendency of younger hiring managers to translate an entire human life into organisational convenience. They speak in KPIs, performance indicators, productivity targets, adaptability, culture fit, and delivery metrics, as though decades of accumulated judgement, survival, disappointment, recovery, reinvention, and hard-won technical fluency exist mainly to drive someone else’s quarterly dashboard. At a certain point in one’s career, the question is no longer whether one can contribute to organisational outcomes. Of course one can. The question is why human experience is so casually reduced to whether it can be extracted, measured, and monetised efficiently enough.
That mindset is profoundly dehumanising. A human life is not reducible to a dashboard metric. Decades spent solving problems, recovering failing projects, stabilising institutions, mentoring others, surviving economic shocks, navigating political environments, and carrying operational accountability cannot be meaningfully collapsed into whether somebody appears energetic enough during a forty-minute panel interview or sufficiently excited about the latest automation trend. There is a coldness in that logic that deserves to be named. It treats human beings less as people with histories and more as units of extractable value.
This is also why I am cautious about the growing expectation that experienced professionals should enthusiastically participate in training the very systems being marketed as their replacements. Increasingly, highly skilled specialists are approached to annotate data, refine outputs, review generated content, or provide domain expertise under the language of “contributing to the future”. I have received paid offers from Micro1 and similar AI-related recruitment or task platforms to participate in that kind of work, and I turn those offers down.
Not because I reject technology. Clearly, I do not. I use AI. I troubleshoot with it. I experiment with it. I recognise its usefulness. However, there is a significant ethical and professional distinction between using a tool to accelerate specialist work and willingly participating in narratives that actively devalue specialist labour while extracting knowledge from it at scale. Some of the current AI economy feels disturbingly extractive. Experienced professionals are increasingly being asked to transfer years of accumulated judgement, pattern recognition, implementation experience, and domain expertise into systems that are then marketed back to organisations as replacements for expensive human labour.
That is the contradiction I cannot unsee. The same market that tells experienced professionals they are becoming too slow, too expensive, too old, or too resistant to change is also quite happy to mine their expertise in order to make automated systems more convincing. It wants the knowledge without the human. It wants the judgement without the person. It wants the pattern recognition without the life that produced it.
The irony is that while these narratives about replacement grow louder, many of us quietly working inside actual systems are seeing something far less dramatic and far more interesting. Recently, I spent hours troubleshooting what appeared to be a tiny issue inside a point-of-sale style interface. A receipt number field was stripping leading zeros. Enter “02”, and the interface reformatted it as “2”. On the surface, it sounded trivial — exactly the sort of thing modern AI marketing would have you believe should be resolved almost instantly by automated tooling.
Naturally, AI became part of the troubleshooting process. And to be fair, it responded quickly and enthusiastically. It proposed JavaScript rewrites, formatting adjustments, string conversions, observer modifications, numerical coercion fixes, and several entirely different implementation pathways. Some of the suggestions sounded convincing. Some partially worked. Others introduced new inconsistencies elsewhere in the logic. Yet none of them truly solved the issue.
What slowly became obvious was that the AI was operating inside the wrong conceptual framework. The system itself had inherited assumptions from an older architecture where the field was originally treated as a currency value. Existing scripts manipulated the value mathematically in the same way a traditional POS keypad handles money input:
val = (val * 10) + pressed;
For monetary calculations, that logic is perfectly valid. However, receipt numbers are not monetary values. They are identifiers, and identifiers frequently require leading zeros. The issue was not really about syntax at all. The issue was that the system no longer understood the meaning of the data it was processing.
The AI continued attempting to repair symptoms while still operating inside the same flawed assumption. At a certain point, the most experienced decision available was not to continue generating fixes, but to stop altogether. That pause changed everything. The next day, the troubleshooting restarted slowly and deliberately. The variables were reviewed line by line. The underlying assumptions were questioned. Existing naming conventions were reconsidered. Then came the most important moment in the entire process: the variable itself was renamed according to its actual purpose.
Instead of continuing to force receipt numbers through a variable associated with monetary input, a new text variable was created and explicitly named “Belegnummer” — identifying it clearly as a receipt or reference number rather than a currency amount. The moment that happened, the entire architecture suddenly made sense again. Receipt numbers are text. Text preserves leading zeros. Therefore stop performing arithmetic on them.
What followed afterwards was straightforward. The keypad logic shifted from mathematical operations to string appending. Display formatting was separated from numerical calculations. The issue disappeared almost immediately. Ironically, this was the point at which AI became genuinely valuable. Once the human specialist established the correct conceptual framework, the AI became extremely effective at accelerating repetitive rewriting, validating patterns, standardising scripts, and reducing manual repetition.
But the critical insight — the reframing of the problem itself — did not emerge from automated generation. It emerged from human judgement. And that is precisely why I remain deeply sceptical of the increasingly aggressive narrative suggesting experienced professionals are somehow becoming obsolete. Much of today’s AI discourse reduces expertise to visible output speed while ignoring the deeper cognitive labour that experienced specialists actually provide.
Real-world systems are rarely clean or isolated. They are messy, layered, political, contradictory, and full of inherited assumptions that only become visible through experience. Experienced professionals are not valuable simply because they can produce deliverables. They are valuable because they recognise when the underlying logic itself is broken. They know when to pause. They know when complexity is becoming performative rather than useful. They know when a system has drifted away from the meaning of the data it was originally built to handle.
That kind of judgement is difficult to compress into a flashy advertisement promising instant transformation through prompts. It is also difficult to assess through the shallow rituals of recruitment panels that behave as though senior professionals must still audition for basic recognition. At this stage in our careers, many of us are no longer interested in performing enthusiasm for systems that extract from people while pretending to empower them.
Technology should support human capability, not become an excuse to erase the value of the humans who built, maintained, repaired, and improved the systems in the first place. AI is powerful. It is useful. It can absolutely accelerate specialist work in remarkable ways. However, there remains a profound difference between accelerating expertise and replacing it entirely.
Personally, I still invest in humans. I will use AI where it is useful, but I am not interested in helping build a professional culture that treats human beings as disposable containers of extractable knowledge. The people most eager to collapse the distinction between assistance and replacement are often those who have never had to carry real accountability for complex systems in the first place.