An interesting take from my studies this week
Here's a statistic that should make every business owner pause: 46% of workers desperately want AI to handle their scheduling, data entry, and repetitive tasks. Meanwhile, 41% of AI startup investment is pouring into areas that workers don't actually want automated.
And on top of that, a 2025 Stanford study found that AI actually fails at boring tasks and thrives at creative work—precisely the inverse of what your team needs.
Welcome to the Tedious Work Paradox. And if you're an Irish SME investing in AI right now, there's a good chance you've got your strategy backwards.
The Misalignment Crisis
These are examples of what is happening in businesses right now. The marketing manager is using ChatGPT to brainstorm campaign ideas and draft social media posts—creative work she actually enjoys. Meanwhile, she's still manually copying data between three different spreadsheets every Monday morning, a task that takes 90 minutes and drives her to distraction.
The operations director is experimenting with AI to "reimagine" your customer journey and "innovate" your service delivery. Brilliant. Except his team is still chasing paper job sheets around the building and re-entering the same customer information into four different systems.
This is the paradox in action. We're automating the meaningful work—the work that requires human judgment, creativity, and satisfaction—whilst leaving the soul-destroying repetitive tasks untouched.
The Stanford research poses a uncomfortable question: "Are we automating meaningful work?" I'd go further: Are we automating the wrong work?
Why This Matters for Irish SMEs
You don't have unlimited resources. Every euro spent on AI implementation needs to deliver tangible ROI, not just impressive demos. Yet I see companies investing in:
- AI chatbots for "enhanced customer engagement" when their CRM is a mess of duplicate records
- Generative AI for "innovative content creation" when their quote-to-invoice process involves printing, signing, scanning, and emailing PDFs
- Machine learning for "predictive analytics" when they can't even reliably track which engineer worked on which job
This isn't innovation. It's distraction wrapped in buzzwords.
The Process Mapping Reality Check
Here's what I believe: before you automate anything, you need to know what you're actually doing. Not what the procedure manual says you're doing. Not what the consultant's flowchart shows. What you're actually doing.
When I work with SMEs, the first thing we do is map current processes—not to admire them, but to expose where the chaos actually lives. And here's what we inevitably find:
The boring work lives in the gaps. It's the data re-entry between systems that don't talk to each other. It's the weekly reconciliation ritual because nobody trusts the numbers. It's the manual scheduling because the diary system predates the smartphone. It's the endless email chains because there's no single source of truth.
This is the 46% that workers want automated. This is where AI could actually save time, reduce errors, and let your people do the work that requires a human brain.
What AI Actually Does Well (Badly)
The jagged frontier of AI capability is critical to understand. AI excels at:
- Pattern recognition across large datasets
- Generating plausible text where "good enough" is acceptable
- Extracting structured data from unstructured sources
- Automating decision trees with clear logic
AI struggles with:
- Tasks requiring precise, deterministic outcomes
- Work that needs perfect consistency
- Processes with frequent exceptions and edge cases
- The boring, repetitive, must-be-right-every-time tasks your team hates
This is the cruel irony. Your marketing manager enjoys writing social posts (creative, variable, good-enough acceptable), so AI can do it. She hates data entry (boring, must-be-perfect, repetitive), but AI struggles with it unless you've structured the process properly first.
The Right Approach: Chaos Signatures First, AI Second
I focus on helping companies move "From Chaos to Control." The chaos comes in recognisable patterns:
Systems chaos: Disconnected processes creating repetitive manual work Financial chaos: Unprofitable operations because you can't track actual costs
Supply chain chaos: Expensive dependencies because you're firefighting, not planning
These chaos signatures are where the boring work breeds. And here's the important bit: you can't AI your way out of process chaos. You'll just create automated chaos, which is worse.
The sequence that actually works:
- Map what you're really doing — Document the current state, warts and all. Where's the re-entry? Where are the manual lookups? Where does information get lost?
- Identify the tedious work — What's boring, repetitive, and adding no value? This is your 46% target.
- Structure the process — Before you automate anything, create the systematic workflow. Clear inputs, defined logic, predictable outputs.
- Then apply AI strategically — Now automation works because you've built the foundation. RAG systems can pull from your organised knowledge base. Process automation can handle your structured workflows. AI can extract data because you've defined what "correct" looks like.
A Practical Example: The Job Sheet Problem
Let me show you what this looks like in practice. I recently worked with a company drowning in paper job sheets. Engineers would go to site, fill out paper forms, come back to office, hand them to admin staff who'd manually enter everything into the system.
The wrong AI approach: "Let's use AI to read the handwritten job sheets!"
Result: 80% accuracy, constant exceptions, admin still checking everything manually, expensive AI solution solving the wrong problem.
The right approach:
- Map the actual job workflow
- Build a mobile-optimised CRM with offline capability
- Engineers enter data on-site with structured dropdowns and validation
- Data flows directly into invoicing, inventory, and scheduling
- AI generates follow-up emails based on job type and findings
Result: No re-entry, no paper, no handwriting recognition needed, engineers happier (easier than paper), admin freed up for actual customer service work.
The AI piece is the smallest part. The systematic process design is what makes it work.
The Questions You Should Be Asking
Before you invest another euro in AI, ask yourself:
Where is the boring work in our organisation? Can you list the top five tasks your team complains about? The ones that are mind-numbing, repetitive, and nobody wants to do?
What are we re-entering? If the same information appears in multiple systems and humans are copying it, that's your target.
Where do we reconcile things manually? Weekly spreadsheet reconciliations are a huge red flag. Why don't the numbers match automatically?
What would our team automate first if they could? Ask them. The answer might surprise you.
What processes do we actually need to fix first? Before automation, before AI, what systematic workflow would eliminate the boring work?
The Bottom Line
The tedious work paradox exists because we're seduced by the impressive demos. AI writing marketing copy is flashy. AI generating creative ideas feels futuristic. AI brainstorming product innovations sounds strategic.
But your competitive advantage doesn't come from AI doing the fun stuff. It comes from systematically eliminating the boring stuff so your people can focus on what actually requires human judgment, creativity, and expertise.
The Stanford study found that workers want help with scheduling, data entry, and repetitive tasks. That's not a coincidence. That's where your chaos lives. That's where time gets wasted. That's where errors breed.
Fix the chaos first. Structure the processes. Then apply AI strategically to handle the boring work your team shouldn't be doing manually.
That's how you move from chaos to control. That's how you get actual ROI from AI investment. And that's how you build a business that works systematically instead of heroically.
Before investing in AI, it's worth taking an honest look at where your time actually goes. That's what a Digital Transformation Assessment does — it shows you where automation will genuinely save hours, not just look impressive on a slide deck. See what's included →
References
Brynjolfsson, E., Li, D. & Raymond, L.R. (2023) Generative AI at work. National Bureau of Economic Research Working Paper Series, No. 31161.
Deloitte (2024) State of generative AI in the enterprise. Deloitte AI Institute. Available at: https://www.deloitte.com (Accessed: 5 February 2026).
Dell'Acqua, F., McFowland III, E., Mollick, E.R., Lifshitz-Assaf, H., Kellogg, K., Rajendran, S., Krayer, L., Candelon, F., Lakhani, K.R. & Bernstein, E. (2023) Navigating the jagged technological frontier: field experimental evidence of the effects of AI on knowledge worker productivity and quality. Harvard Business School Technology & Operations Management Unit Working Paper No. 24-013.
Eurostat (2025) AI adoption in European enterprises. Statistical Office of the European Union. Available at: https://ec.europa.eu/eurostat (Accessed: 5 February 2026).
International Labour Organization & United Nations (2024) Mind the AI divide: shaping a global perspective on the future of work. Geneva: ILO Publications.
Lee, H.P., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R. & Wilson, N. (2025) 'The impact of generative AI on critical thinking: self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers', in Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems. New York: ACM, pp. 1–22.
Morandini, S., Fraboni, F., De Angelis, M., Puzzo, G., Giusino, D. & Pietrantoni, L. (2023) 'The impact of artificial intelligence on workers' skills: upskilling and reskilling in organisations', Informing Science: The International Journal of an Emerging Transdiscipline, 26, pp. 39–68.
Noy, S. & Zhang, W. (2023) Experimental evidence on the productivity effects of generative artificial intelligence. MIT Economics Working Paper.
Postman, N. (1998) Five things we need to know about technological change. Talk delivered at the New Tech '98 Conference, Denver, Colorado, 28 March.
Toner-Rodgers, A. (2024) Artificial intelligence, scientific discovery, and product innovation. arXiv preprint arXiv:2412.17866. Available at: https://arxiv.org/abs/2412.17866 (Accessed: 5 February 2026).
Note on Stanford Study (2025): The statistics regarding worker preferences for AI automation (46% wanting help with tedious tasks vs. 41% of investment targeting other areas) are cited from recent Stanford research as presented in Pietrantoni, L. (2025) Integration of AI in organizations [Lecture presentation]. University of Bologna, 5 February.