Help Someone With AI: Solve One Real Problem
Help someone with AI by guiding one real, contained work problem. In this episode, you will learn how to choose a safe text-heavy task, set clear boundaries, guide the prompt and capture a useful outcome without becoming the unofficial AI helpdesk.
Start with the Episode 16 podcast.
Listen for the difference between doing someone else’s work for them and guiding a short, structured AI peer consult that creates useful proof of capability.
Watch the AI peer consult sequence.
Use the video to see how a small, contained consult can isolate an admin bottleneck, set boundaries and keep the work owned by the person you are helping.
AI capability becomes more credible when someone else can explain what improved.
To help someone with AI well, keep the task defined, safe and owned by them. The aim is not to become the unofficial AI helpdesk. It is to guide one useful process and capture a simple piece of external evidence.
How to help someone with AI through a short peer consult.
When you help someone with AI, your private capability becomes visible through a clear work outcome. Episode 16 moves your AI use out of your own workspace and into a small, useful collaboration with another person.
The goal is not to take over their task. The goal is to guide a safe, focused session where they keep ownership of the work, understand what is happening and can explain what improved. This practical approach connects with the wider Season 2 AI podcast and the hands-on Zero to AI Learning Labs.
The simple peer-help loop
A good AI peer consult stays small, practical and safe. You are not redesigning someone else’s whole workflow. You are helping them solve one clear bottleneck, then recording what changed.
Find the bottleneck
Look for a colleague who regularly loses time to summaries, updates, analysis notes, briefing material or another repetitive text-heavy task.
Keep the scope small
Choose one task that can be improved safely inside a short working session, without taking responsibility away from the person who owns it.
Guide the prompt work
Talk through the prompt structure, context, source material and checks while your colleague operates the AI tool on their own screen.
Capture the result
Record what the work was like before, what changed afterwards and a short written reference from the person you helped.
What to look for
Good candidates are repetitive, text-heavy tasks that are easy to explain, easy to check and low enough risk to practise on safely. Start with work where the person can compare the AI-assisted result with the original process.
- Summaries Meeting summaries, consultation notes, progress updates or short research digests.
- Briefing material First-pass briefing notes, stakeholder updates or structured decision summaries.
- Document clean-up Restructuring rough notes, turning bullets into clearer prose or improving readability.
- Admin bottlenecks Repetitive text tasks that drain time but do not require sensitive or high-risk data.
Set clear boundaries
A useful consult should protect trust. Keep the work on your colleague’s screen, let them control the keyboard and avoid moving their material into your own private account. Before using personal information, review the New Zealand Office of the Privacy Commissioner’s guidance on artificial intelligence and privacy. For wider risk, transparency and accountability checks, use the official Responsible AI Guidance for the Public Service.
- Keep ownership with them They should see, check and own the output.
- Stay inside a safe task Avoid confidential, sensitive or high-risk work unless organisational rules clearly allow the tool, data and process.
- Check the original material If source material is involved, manually verify important claims, references and judgement calls.
- Ask for a short reference A useful reference can be simple: what was hard before, what changed and whether the workflow helped.
Help someone with AI on one text-heavy task.
Find a colleague with a small, safe bottleneck. Keep the session contained, work on their screen, help them structure the prompt, check the result and write one short note about what improved.
Use this peer consult prompt as a starting point.
This is not a form and nothing is saved on this page. Copy the prompt manually into your preferred AI tool, or explore more practical options in the AI learning resources and tools.
Act as a practical AI consult coach for an experienced professional. Help me plan how to help someone with AI through a short, safe session focused on one text-heavy task. Ask me who I am helping, what task slows them down, what source material is involved, what risks or privacy issues I need to avoid, what prompt structure would help, how I can keep the work owned by them, what checks are needed, and what simple before-and-after evidence I should capture.
Help someone with AI and make your capability visible.
Choose one safe peer problem, guide the process carefully and keep a short record of what improved. Continue with the next episode to turn your own working patterns into a practical AI doppelganger workflow.