Try OnsaEvery few weeks, someone publishes a list of jobs that AI will “eliminate.” Data entry clerks. Paralegals. Junior developers. Customer service reps.
These lists get clicks, but they’re asking the wrong question.
The better question: which tasks within a job get automated? And are those the expert tasks or the routine ones?
The answer determines whether wages go up or down.
A paper from Digitalist Papers Vol. 2 laid this out clearly. The authors analyzed US labor market data from 1980 to 2018 and posed a puzzle:
Why did the same technology—computers—lead to completely different outcomes for accountants and warehouse workers?
Accountants got raises. Warehouse workers got pay cuts.
Same technology. Opposite results.
Their answer came down to what got automated.
For accountants: Computers automated routine, non-expert tasks. Data entry. Calculations. The boring stuff that anyone with training could do.
For warehouse workers: Computers automated their most expert tasks. Knowing where things are stored. Navigating the layout efficiently. The knowledge that took years to build.
When expert tasks get automated, barriers to entry drop. More people can do the job. Wages fall.
When non-expert tasks get automated, the opposite happens. The expert work becomes more valuable. Wages rise.
At Onsa, we automate the admin work in sales: lead sourcing, research, qualification, data entry, initial outreach drafts.
That’s routine work. It’s important, but it’s not the expertise that closes deals.
The expertise in sales is relationships, politics, reading the room, selling the vision, navigating complex organizations. That doesn’t get automated—it becomes more valuable.
My prediction: entry-level sales salaries will compress. The “research and prep” work that justified junior headcount gets automated. But top performers who can close complex deals? They’ll command premiums.
Sales cycles will compress too. The endless discovery call → another call → meet the senior person → proposal → negotiation theater will fade. Faster qualification means faster decisions.
This one’s already visible.
Junior developer hiring is feeling the squeeze. AI can write boilerplate code, handle routine tasks, and accelerate simple implementations. The work that used to train junior devs is increasingly automated.
Meanwhile, AI researchers and senior architects command astronomical salaries. The expert work—system design, architectural decisions, debugging novel problems—is more valuable than ever.
Same pattern: routine tasks automated, expert work appreciates.
Paralegals doing document review? That’s getting automated fast.
Partners who understand client politics, structure complex deals, and provide strategic counsel? Their value increases.
Junior associates are in the awkward middle. Some of their work is automatable (research, drafting). Some isn’t (client relationships, judgment calls). The role will evolve, not disappear.
If you’re worried about AI taking your job, ask yourself:
What tasks do I spend time on?
Make a list. Literally write them down.
Which ones are expert tasks vs. routine tasks?
Expert tasks require judgment, relationships, or knowledge that takes years to build. Routine tasks are important but teachable—anyone with training could do them.
Which are getting automated?
If AI is automating your routine tasks, that’s probably good for you. More time for expert work.
If AI is automating your expert tasks—the things that made you valuable—that’s a warning sign.
“AI will take jobs” is too simple.
AI automates tasks. Which tasks get automated determines whether you benefit or suffer.
The strategic question isn’t whether to use AI. It’s which tasks you choose to automate—and whether you’re building expertise in the parts that remain human.
For companies: be thoughtful about what you automate. It’s not just a technology decision. It’s a decision about what kind of jobs you’re creating and what skills you’re valuing.
For individuals: pay attention to where the value is moving. If AI handles your routine work, lean into your expert work. If AI is coming for your expertise, start building new expertise.
The future isn’t humans vs. AI. It’s humans with AI—and the question is which humans are amplified and which are displaced.
I’m Bayram, founder of Onsa. We automate the routine parts of sales prospecting—research, qualification, outreach—so salespeople can focus on relationships and closing. If you want to talk about where AI fits in your sales process, find me on LinkedIn.
P.S. The Digitalist Papers research is worth reading if you want to go deeper on the economics.