The workplace has embraced AI at remarkable speed, with many professionals now using it to write, research, summarise meetings and streamline routine tasks. But as AI becomes a regular part of the workday, a key question remains: is it improving how work gets done or simply adding more activity to the day?
“Using a tool daily does not necessarily mean you're being productive,” says Ziyanda Gxuma, Scrum Master at Strider Digital. “It's easy to measure frequency by counting how many prompts you enter each day but productivity is much harder to measure.”
For Gxuma, the real value of AI lies in removing repetitive work so that more time can be spent on collaboration and the kinds of work that require deep thinking and experience.
Saving time isn't the same as being productive
One of the most common assumptions about AI is that it automatically reduces workload. In practice, getting useful output often requires iteration and refinement.
“People who’ve been using AI tools more and more frequently quickly realise that outputs need to be corrected often and prompts need to improve as you work towards the right result,” says Gxuma, “and this process can take minutes or even hours, meaning it sometimes takes longer than just completing the task on your own.”
She also notes that AI output requires scrutiny rather than mere acceptance.
AI can easily misunderstand prompts or provide inaccurate information, it's one’s own knowledge, experience and understanding of the context that enable AI to produce the best possible responses.
Treat AI as a collaborator, not a replacement
Rather than seeing AI as a shortcut to answers, Gxuma sees it as a support tool that strengthens thinking and review processes.
“I use AI as a research assistant and sometimes as a critic of work I've already produced to help identify gaps I may have not considered,” she explains. “I've also found that the best use of these tools isn’t asking for answers but rather learning how to ask better questions.”
That shift introduces its own complexity, however. Crafting prompts, reviewing responses and refining them can take time and sometimes interrupts workflow. But she argues this process is very necessary because accepting AI-generated content simply because it sounds convincing can be highly risky.
Context, she adds, is especially important in collaborative environments. If AI is asked, for example, to generate improvement suggestions for a team without enough context or without careful review, it can produce recommendations that are difficult to justify.
Critical thinking remains essential
As AI becomes more capable, it still relies heavily on the quality of human input and interpretation. Gxuma says this makes evaluation and context-setting a core part of using these tools well.
“Again, you need to evaluate the accuracy of the information AI provides, consider your own context and challenge any assumptions it has made,” she says. “AI isn't here to replace human thinking, it's here to enhance it by acting as a sounding board and creative partner in a human-led conversation.”
In practice, that means the usefulness of AI output depends on how much relevant detail is provided upfront. Without it, even well-structured responses can miss the mark or produce ideas that don’t fit the situation. Gxuma notes that the goal is not just to get answers but to generate something that can be developed further.
Don’t let AI replace your expertise
While much of the conversation around AI focuses on job displacement, Gxuma says a more immediate risk is the gradual weakening of core skills when they are not actively used.
“Heavy reliance on AI can weaken skills like writing, analysis and problem-solving if we stop practising them ourselves but interestingly it can also have the opposite effect if used properly,” she says.
She points to her own experience with sprint goals, a task she initially struggled with and began using AI to support. Over time, she noticed something unexpected: instead of replacing her thinking, the process re-engaged skills she had previously set aside from her earlier career as a writer. Paying closer attention to AI outputs forced her to revisit language, structure and tone more deliberately.
Good judgement will set AI users apart
As AI becomes more integrated into daily work, the difference between effective and ineffective use comes down to how intentionally it is applied and whether outputs are reviewed, contextualised and combined with human input rather than taken at face value.
“The differentiator, in the end, will be good judgement. The strongest users will know when to use AI, when not to, how to verify outputs, how to provide context and how to combine AI assistance with human expertise,” concludes Gxuma.
