Maybe We’re Asking AI to Do the Wrong Work

Maybe We’re Asking AI to Do the Wrong Work

Max Symuleski

Max Symuleski

September 30, 2026

More output does not necessarily mean better work. In fact, when AI allows people to create at a speed they could never sustain on their own, it can become harder to give every output the same attention and care.

I say this as an early AI adopter who uses these tools every day. Some of my favorite apps are also the least glamorous. I use AI to organize research sources, structure information in spreadsheets, and turn raw data from lengthy studies into something people can actually read.

None of that sounds as impressive as asking AI to write a paper or generate a campaign. But it gives me more time to read, interpret and think critically about the research myself.

We have spent a lot of time asking what AI can do for people. We should spend more time asking what people actually want it to do for them.

Do people want AI doing interesting work?

Because generative AI can produce text, images and ideas in seconds, companies have focused on using it to automate creative and knowledge work. Yet there is a meaningful difference between a task AI can perform and one a person wants to delegate.

Many people derive purpose, satisfaction and even enjoyment from the parts of work that require judgment, creativity or close attention. Automating those tasks may increase output without improving the experience of doing the work, or the quality of what gets produced.

The goal should not always be getting AI to do more of our work. Sometimes, it should be getting the boring work out of the way so people can spend more time on the parts that require uniquely human attention.

That may mean organizing the research rather than interpreting it, or formatting the data rather than drawing conclusions from it. The value comes from creating space for people to do more of what they find meaningful, not simply producing more.

What do people actually want AI to do for them?

AnswerLab is currently exploring this question through new research into what people are willing to hand over to AI, what they want to remain in control of and why. I will not get ahead of the findings, but the premise raises an important question for every organization: Are we using AI for the work people genuinely want help with, or simply the work the technology can already do?

In the workplace, some of the most valuable applications may also be the least glamorous: cleaning up data, formatting documents, organizing files or structuring information. These tasks rarely receive the most attention, but automating them can give employees more time for analysis, problem-solving and creative thinking.

That matters because AI does not always eliminate work. Sometimes it simply moves it. The more content a model produces, the more time someone must spend reviewing, verifying and correcting it. An application that appears to save time on the front end may quietly reclaim that time through additional oversight.

If that verification work is going to exist either way, it should be designed into the experience—not left for users to manage on their own. Some tools have begun incorporating source links and connecting outputs to specific supporting material, but no one has solved this perfectly. The products that make it faster and easier to understand what AI did, why it produced a particular answer and where the information came from will create more meaningful value than those that treat verification as the user’s problem. 

The better question is not how much work AI can perform, but which parts of the work it should own. The most technically impressive application is not automatically the most valuable one.

Build AI around human value, not technical capability

Most AI product strategies still start from the same place: the technology can do this, so where do we apply it? 

That question has produced a lot of impressive demos and a lot of features people try once and never open again. The more interesting question runs in the other direction.

Start with the friction people actually want removed, the tasks they would happily never do again, and build from there. But look beyond the individual task to the workflow around it. AI output is rarely the end of anything; it typically lands in the middle of a process where someone must make sense of it, connect it to what they already know and decide what happens next. The most valuable products will support that entire workflow, giving people time and attention back for the work and experiences that matter to them.

That is where AI creates real value, not by doing more of our thinking, but by making more room for it.

About the author

Max Symuleski

Max Symuleski

AI Product Manager and Principal AI & Research Innovation Strategist

Max Symuleski, Ph.D., is the Principal AI & Research Innovation Strategist at AnswerLab, where they lead strategic initiatives that integrate AI and emerging technologies into the company’s qualitative research practice and future experience strategy services. They specialize in foundational and evaluative research around emerging tech and AI, with deep expertise spanning ranking and recommender systems, generative AI, and responsible/ethical AI. In their current role, Max helps shape the vision and roadmap for AnswerLab’s AI-enabled research capabilities—co-developing internal tools, AI-assisted workflows, prompt frameworks, and reusable assets that scale high-quality qualitative research while preserving rigor and nuance. They partner closely with researchers, project managers, and strategists to translate research and business needs into AI-enabled solutions, champion internal adoption, and ensure every innovation aligns with AnswerLab’s values around research quality, ethics, and strategic impact. Max also builds and supports AnswerLab’s internal AI SME network, enabling distributed experimentation, training, and AI fluency across teams. Beyond internal innovation, Max acts as a strategic thought partner to clients whose products, features, or audiences intersect with AI. They advise on research designs and AI-related considerations, contribute to thought leadership, and help organizations navigate emerging AI trends, risks, and opportunities. Max holds a Ph.D. in Computational Media, Arts, and Cultures from Duke University.

Connect on LinkedIn

Newsletter

Insights directly to your inbox

  • Trusted by product and digital leaders at future-thinking brands
  • Curated articles, case studies, and impactful resources
  • No spam, unsubscribe anytime