Digital fluency: why skills beat borrowed knowledge
Skills over knowledge
Knowing about AI is not the same as using it when the pressure is on. Knowledge you have not applied is fragile. The first time someone challenges you, you end up defending something you only skim-read.
Practised knowledge is different. When you have tried, iterated, and learned, you can explain what worked, what failed, and why you chose a path. Read widely, then adapt what you learn, test it in your work, and make it your own.
Aim to be a learn-it-all, not a know-it-all. That mindset pushes you to seek opportunities to learn every day, not look for ways to skip the practice.
Why digital fluency matters now
For years, many organisations treated upskilling as an occasional event, not a continuous habit. People filled gaps informally, asked the person next to them, or searched when stuck. It worked, until AI became ubiquitous at work and sped up change.
Today, tools change monthly. Threats scale with them. We haven’t invested in training for the last two decades and so can’t expect to give teams hours of self-paced video training as way to ‘binge-watch’ their way out of the digital skills gap. Instead, we should treat digital fluency like fitness and health. Show up every day, tweak and adapt based on results.
Let’s look at a definition: Digital fluency is applied, confident use of digital tools to achieve outcomes, safely and ethically.
Fluency sits on four pillars:
Information and data sensemaking. Understand your data estate, the context of requests, and the provenance of sources.
Problem solving. Use systematic approaches rather than impulsive reactions.
Collaboration. Get a second set of eyes before publishing anything that could expose people or organisations.
Safety and ethics. Strengthen security behaviours, and don’t stop at “can we” but also ask the “should we”.
From change fatigue to curiosity
Not everyone enjoys change. Start where you are. Turn big changes into small, repeatable actions.
A simple way to diagnose why you feel stuck:
Meaning: do you understand why the change matters to you and your team?
Know-how: do you know what good looks like, and where to learn it?
Ability: have you practised enough in a low-risk setting?
Momentum: do you have a small win to reinforce the habit?
Celebrate your progress. The thing about consistent practice is you forget how far you’ve come. Think back to where you started, and keep an eye on where you’re heading.
Trust, security, and AI-scaled threats
Security questions like “first pet’s name” belong in a museum. Instead use passkeys and multi-factor authentication (MFA) that uses an app prompt or security key. Reduce what can be phished, reused, or guessed.
Adopt Zero Trust habits in everyday work:
Verify explicitly. Do not rely on location, device type, or job title to grant access.
Limit access. Give people the least they need for the task. Remove access once the task ends.
Assume breach. Log, alert, and review. Treat anomalies as early warnings, not noise.
Two quick wins that you can take this week:
Aim to turn MFA on everywhere. Prioritise accounts that control money, identity, or admin rights.
Run a pre-post check. Screenshots of whiteboards, Miro boards, and dashboards often carry names, emails, IDs or other personal information you might not intend to share.
Consumer platform policies and wearable partnerships change fast. If you operate public profiles for work, review privacy settings, your posting habits, and what you consent to. Decide what data leaves your device, who processes it, and what happens if ownership or partners change. Build a simple checklist so you are not surprised later.
Go one level deeper
Make “one level deeper” your weekly ritual. If you meet a new acronym, ask your AI assistant to explain it in practical terms. If you rely on a model output, ask it to cite and compare sources, then spot-check them yourself.
Three areas where one level deeper pays off fast:
LLMs (large language models). Learn about grounding, hallucinations, and why better prompts are not a substitute for source verification.
Identity. Learn passkeys, conditional access, device compliance, and trust.
Content provenance. Explore watermarking, content credentials, and how to spot weak signals of synthetic media.
Try this prompt for writing proposals
This is a good one for Microsoft 365 Copilot which has access to your organisational data like emails in Outlook, files in Sharepoint, and messages in Teams. Provide context on the project, audience, sections, word counts, tone, and any non-negotiables requested in the Request for Proposal (RFP) or Request for Information (RFI).
Prompt for one section at a time. Specify what to include, what to avoid, and how you will judge success.
Example section prompt
You are writing the “Problem we solve” section for a proposal.
Audience: senior business leaders in New Zealand banking.
Length: 180–220 words.
Tone: concise, confident, plain English, NZ spelling.
Include: the cost of delay in qualitative terms, a short scenario, and one measurable outcome.
Exclude: buzzwords, clichés, promises without evidence.
Use information from the attached files where relevant. If uncertain, ask for one clarifying detail.
Export your instructions/prompts as JSON (JavaScript Object Notation) so you can reuse and adapt them later.
Use text-to-speech to hear the draft, or read it to a friend or colleague. Your ears catch clunky phrasing your eyes ignore.
A weekly practice plan
Monday: Replace one weak login with a passkey.
Tuesday: Run a 10-minute privacy sweep on one app you use for work.
Wednesday: Practice one research prompt with clear success criteria.
Thursday: Draft or refactor one section of a work document with AI. Save your prompt context.
Friday: Teach a colleague one skill you learned. Teaching locks knowledge in. Ask them to teach you one too.
Two everyday AI use cases
Travel disruptions. When flights change, ask AI to summarise your options, your rights, and the fastest contact path. Provide dates, booking numbers, traveller status, and your goal. Ask for a summary script to use on your call so you hit the most important points to get the support you need.
Subscription control. When signing up for a trial, ask for exact steps to cancel before renewal. Save the steps with a calendar reminder. Decide in the clear, not under auto-renew pressure.
The takeaway
Digital fluency is the ability to apply, adapt, and explain what you are doing, safely and ethically, with modern tools. Practice small, secure habits, and finish each week with technical expertise that’s one level deeper than where you started.
