The Job Did Not Disappear. It Changed Shape

Every few years somebody announces that machines are coming for our jobs, and every few years the prediction lands somewhere between wrong and beside the point. What actually happens is quieter and harder to argue with. The job title stays. The work inside it changes. And the person who does not notice the change is the one who gets caught out.

We see this up close. Lightwins Creations builds systems for microfinance institutions, churches, schools, NGOs and SMEs across Zambia, which means we spend a lot of time watching what happens to the people whose daily work our software touches. The pattern is consistent enough to be worth writing down.

Three shifts that are already here

Tools got cheap. Capabilities that used to require a budget and a specialist now cost the price of a monthly subscription. Design, video editing, transcription, translation, first-draft writing, basic data analysis. This does not automatically eliminate the specialist. It does mean the specialist is now competing with a good-enough alternative that costs almost nothing, and that changes what they have to be good at.

The work moved up a level. When a task gets automated, the human job usually shifts from doing the task to checking, framing and owning it. A loan officer who used to spend the morning writing up repayment schedules by hand now spends it deciding which cases are actually risky. That is a harder job, not an easier one, and it demands different skills.

Proof moved online. A decade ago your reputation lived in the people who knew you. It still does, but now it also lives in what a client finds when they search your name before the meeting. That cuts both ways. It is the cheapest professional advantage available to anyone willing to build it, and the most obvious gap when it is missing.

An office desk and computer, representing roles being reshaped by new technology
The desk stays the same. What happens at it does not. Photo by Alesia Kazantceva on Unsplash.

What this looks like in Zambian workplaces

Loan officers and credit staff. Portfolio reports, ageing analysis and arrears listings that used to take a day now come out of the system in seconds. The job is moving toward client judgement, field relationships and knowing which borrower is telling you the whole story. The officers who thrive are the ones who read a report critically rather than simply producing one.

Bookkeepers and accounts staff. Data entry is shrinking. Reconciliation, compliance and interpretation are not. With ZRA systems tightening and reporting standards getting stricter, the person who understands why the numbers say what they say is more valuable, not less. The person who only keys them in is exposed.

Media and communications teams. This is the sharpest change. Editing, graphics, voiceover and basic video no longer require a large team or an expensive suite. But the value has shifted to the things the tools cannot supply: knowing what the audience actually responds to, showing up reliably, and understanding the institution well enough to speak for it. A church media volunteer who understands the congregation is worth more than a technically better editor who does not.

Administrators and secretarial staff. Scheduling, minute-taking, form filling and routine correspondence are being absorbed by software. Coordination, discretion and institutional memory are not. The distinction matters, because most administrative roles contain a great deal of both, and the second part is rarely the part written in the job description.

Teachers and trainers. Content is now free and abundant. Attention, discipline, assessment and the ability to tell whether a pupil actually understood something remain stubbornly human. Schools that treat technology as a replacement for teaching tend to regret it. Schools that treat it as a way to free the teacher from admin usually do well.

What is genuinely hard to automate

It is worth being precise here, because vague reassurance helps nobody. Four things have held up consistently:

  • Accountability. Someone has to sign, and be answerable for it. Software does not carry consequences.
  • Relationships built over time. Trust with a client, a congregation, a board or a donor is slow to build and cannot be generated on demand.
  • Judgement in messy situations. The cases that do not fit the policy are exactly the ones that need a person.
  • Physical presence. Field visits, installations, service delivery, being in the room.

Notice what these have in common. None of them are technical skills. The most durable parts of most jobs are the parts that were never really about the tools in the first place.

Colleagues working together around a table
The parts of the job that survive are usually the parts that involve other people. Photo by Annie Spratt on Unsplash.

What to do in the next twelve months

If you are employed, four things are worth doing, roughly in this order.

  1. Learn the tool that touches your own work. Not technology in general. The specific system your organisation uses, properly, including the parts nobody trained you on. Being the person who actually understands the system is a real position to hold.
  2. Get comfortable with AI tools rather than avoiding them. Spend a few hours using them for real tasks. The gap opening up is not between people and machines. It is between colleagues who use the tools well and colleagues who do not.
  3. Move one step toward judgement. Whatever part of your job involves deciding rather than processing, ask for more of it. That is where the role is heading anyway.
  4. Make your work findable. A tidy LinkedIn profile, a proper email address, a record of what you have delivered. Low effort, disproportionate return.

If you employ people

The temptation when a system automates a function is to cut the post. Sometimes that is correct. More often it is expensive, because you lose the institutional knowledge and then discover the system needed a person to run it properly all along.

The better question is what the freed-up hours should now be spent on. In most of the institutions we work with, there is a long list of things nobody has had time to do: client follow-up, arrears recovery, donor reporting, member engagement, data cleanup. Redeploying is almost always cheaper than replacing, and staff who were part of the change tend to defend it rather than quietly work around it.

Budget for training as part of the system cost, not as an afterthought. A system nobody was properly taught to use is not a system. It is a licence fee.

The uncomfortable part

None of this is evenly distributed. Someone with reliable internet, a decent device and an employer who pays for training will adapt far more easily than someone without. That is a real inequality and pretending otherwise helps no one. But the entry cost is lower than it has ever been, and the skills that matter most are not the expensive ones. Curiosity and a willingness to be a beginner again go a long way.

The jobs are not disappearing. They are being rewritten, and mostly the rewrite favours people who can think, decide and be relied upon. That has always been the case. Technology has just made it more obvious.


Lightwins Creations builds and supports technology for Zambian microfinance institutions, churches, schools, NGOs and SMEs, and we train the teams who use it. If you are working out what a new system means for your staff, or which roles need to change and how, talk to us.

Woodgate House, Cairo Road, Lusaka
[email protected] | +260 976 351 166

Posted in Ai Technology

Leave a Comment

Your email address will not be published. Required fields are marked *

*
*