Artificial intelligence (AI) is revolutionising education. AI can automate grading, generate lesson plans and provide instant feedback to students. But for many educators, it has not reduced workload; it has reshaped it, often in ways that feel just as overwhelming as before. The temptation to adopt multiple AI tools without a clear strategy has left many teachers spending more time learning new technology than actually teaching.
We have found that the key to making AI work for educators is not using more AI but using it more deliberately. Here are four ways to help educators integrate it into their teaching in a way that supports, rather than overwhelms.
Disrupt the routine
Repetitive teaching tasks, such as grading similar assignments, delivering the same lectures and writing identical pieces of feedback, can accelerate burnout. While AI can streamline workflows, over-reliance on a single function risks creating new forms of monotony. Strategic dishabituation, which involves intentionally disrupting predictability, helps keep teaching engaging without adding to the tech burden.
Beyond these structural shifts, we varied our AI techniques, exploring tools to facilitate interactive exercises, grammar feedback and dynamic assessments rather than using AI solely for lesson planning. Additionally, random AI-generated warm-ups and spontaneous review sessions inject variety into lessons, breaking rigid teaching patterns.
By regularly switching thematic focus, for example, experimenting with new assessment methods one term and personalised learning the next, we ensure that AI-enhanced teaching remains stimulating rather than routine. Small, intentional adjustments can transform AI use from monotonous to innovative.
Focus on using AI for workflow optimisation
AI adoption is often marketed as a way to save time. However, without a proper strategy in place, it can create as much work as it removes. We realised that adopting AI in a way that optimises our workflow requires being selective about how we use it.
Rather than automating everything, we focus on a few high-impact areas:
Automated grading: AI provides initial feedback on formative review assessments, allowing us to focus on student support and other teaching activities.
Lesson structuring: instead of writing every lecture outline from scratch, AI helps generate structured templates that we can refine.
Administrative tasks: AI-assisted email drafting cuts down the time spent on responding to student queries and taking meeting notes.
We avoid relying on AI for summative assessment feedback, personalised student feedback, lecture delivery and complex academic discussions: areas where human expertise is irreplaceable.
- Bringing GenAI into the university classroom
- An insider’s guide to how students use GenAI tools
- So, you want to use ChatGPT in the classroom this semester?
Adopt a minimalist approach
Just because an AI tool exists, it doesn’t mean it’s useful. Many educators fall into the trap of AI overload, testing multiple tools and adding unnecessary layers of technology to their workflow. Instead, we took a minimalist approach, using an AI streamlining checklist before adopting any new tool. Questions to ask yourself include:
- Does it genuinely save time?
- Does it integrate easily into your workflow?
- Does it simplify rather than complicate your work?
Rather than juggling multiple platforms, we focus on a set of AI tools that align with our teaching needs. For instance, Eduaide.Ai provides ready-made lesson plans, quizzes and learning activities, reducing preparation time without requiring extensive manual input. This allows us to maximise efficiency while keeping AI integration simple and manageable.
By using only two or three high-impact AI tools, we avoid the mental fatigue of constantly switching between platforms. AI should be an assistant, not a distraction.
Use for feedback and self-reflection
Many educators use AI to provide feedback to students, but it can also support reflection on teaching practice.
We use Otter.ai to transcribe our lectures and review them afterwards, helping us to identify where student engagement dropped, where explanations could be clearer, or where interaction was lacking. Similar insights can now be gathered using simple voice recordings on our phones, which can be transcribed and analysed with AI tools like ChatGPT.
We also use ChatGPT to review samples of our written feedback, identifying patterns in tone and content. This helps us vary our language and ensure greater clarity and consistency across student comments. Additionally, RescueTime allows us to track how much time we spent on grading, planning and admin tasks, helping us to pinpoint where we can streamline our workload.
Together, these tools support small but meaningful improvements in both our teaching delivery and time management.
Isaac Asimov’s warning that “The saddest aspect of life right now is that science gathers knowledge faster than society gathers wisdom” rings true in an age that is becoming dominated by AI use. And AI in education is no different. More tools do not necessarily mean better teaching. Smarter use of fewer tools does.
By rotating tasks you use AI for, optimising workflow, applying a minimalist approach and using AI tools for self-reflection, we have found that these tools can be powerful allies; but only when they serve our teaching goals.
If AI is creating more work instead of reducing it, it might be time to ask: What’s one AI tool or habit you could refine this week to truly make your teaching more efficient?
Garth Elzerman and Bin Feng are educators at Xi’an Jiaotong-Liverpool University in China.
If you would like advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.
comment