AI as Ally: How to Act Early, Learn Fast, and Stay Human in a Machine World

Michael Reid OAM Artist Credit: Adam Cullen

There has been a tendency to scuttle across the arrival of Artificial Intelligence with some anxiety. Deep breath. In responding to AI, I begin with two simple ideas.

The Chinese philosopher Confucius (551–479 BCE) taught that progress is not made from nothing, but through the careful refinement and recombination of what already exists. He is often credited with the sentiment that “there is nothing new under the sun”, although this Biblical phrase is drawn from Ecclesiastes. Both, however, express a similar insight: that innovation is rarely invention, but recognition. A realisation in this case that AI is part of a continuum of human tool making. If this is true, then the tools we need to respond to AI are already at hand. The task is not to wait for wisdom, but to find and recognise where it already lives.

All wisdom comes from memory. If we look back, even briefly, we may find that the framework for engaging with AI is already there, waiting to be recognised. Health educators have long understood that early action during a pandemic—and some would view AI as a pandemic—is far more decisive than later intervention. Swift measures curb spread and save lives, while delay allows for greater escalation. Applied to AI, the lesson is direct. We must act early, learn quickly and embed new systems before opportunity overruns us.

When it comes to engagement, start with your immediate world. Of course, understanding the big picture—even the ecosystem from which AI is emerging—has its place. Reading the Financial Times and absorbing the implications of AI’s future energy demands helps us grasp how much electricity, and how much often-overlooked water, will be diverted to support it. Useful knowledge, but not the crucial first step. Start, as I do, with your own environment. Let AI help you where you are and learn outward from there. Like a stone rippling waves across a pond, start with one splash and all will flow.

Michael Reid OAM , seen here at his gallery in Murrurundi, says AI isn’t the threat … Inaction is as he explains here in his Personal Playbook for the Future

How We Will Use AI

1. General Questions & Ideation → Perplexity & ChatGPT/Claude

  • Perplexity: Quick fact checks, summaries, and context gathering. Use when we need a broad, sourced overview (e.g., “Who represents this artist in Berlin?”).
  • ChatGPT/Claude for Ideation: Use in “brainstorming mode” for creative, divergent thinking. Prompt for “20 unconventional angles for an exhibition on [theme]” or “potential risks and opportunities I haven’t considered.” This uses AI as a tireless, non-judgmental idea catalyst before the critical polishing phase.

2. General Searches & Deep Analysis → DeepSeek

  • Discovery, exploration, and deep analysis of our own documents.
  • Useful for lists, contacts, comparisons, and scanning multiple angles. Strong for data mining and sourcing.
  • Key Feature: Use its ability to process uploaded files (PDFs, Word, Excel) and perform live web searches to analyse our materials in the context of current information.

3. Drafting, Editing & Reviews → ChatGPT

  • Polishing and tightening language: spelling, grammar, tone, and flow.
  • Cataloguing consistency, exhibition texts, board papers, and proposals.
  • Acts as the penultimate review, ensuring clarity and logic before final review.
  • Critical Reminder: Be especially sceptical of AI-generated content that is confidently and eloquently wrong. Always verify core assertions, even—or especially—when they are well-written. The fluency can mask factual or logical fallacies.

4. Email Triage & Drafting → Gemini (in Gmail)

  • First-pass drafting of routine or repetitive emails.
  • Summarising long email threads quickly and accurately.
  • Identifying key actions, deadlines, and unanswered questions.
  • Drafting courteous, clear responses for later review.

Gemini is for speed inside the inbox, not authority. Final emails—especially external, sensitive, or strategic—remain human-written and human-approved.

Practical Workflow

  • For creative projects, begin with a ChatGPT/Claude brainstorming session to generate raw ideas and angles.
  • Run letters, proposals, reports, and PDFs through ChatGPT for polish and clarity.
  • If relevant, run the same document through DeepSeek for factchecking, broader scanning, or deep file-based analysis.
  • Use Gemini in Gmail to summarise complex threads, draft first responses, and surface key actions.
  • Always finish by running everything through Microsoft Spell Check to return the document to Australian English.

Over the past week I have been uploading current artist and exhibition PDFs into ChatGPT and DeepSeek. AI quickly identifies spelling, grammar, and formatting errors that are easy for humans to miss. Remember: “Garbage In, Gospel Out.” AI treats your input as authoritative, so the quality of its output depends on the quality of your source material. Mistakes happen; AI gives us the capacity to improve rapidly.

Final Review & Responsibility

  • Everything written and sent in your name must still be reviewed by you.
  • AI is powerful, but it is not better than us in our chosen fields.
  • Facts can be wrong. You must recognise when they are.
  • Always cross-check important information by running it through multiple AI platforms and tracing back to primary sources where possible.

Guardrails

1. Confidentiality

  • Do not upload highly sensitive legal, financial, or personal documents without approval.
  • Assume anything uploaded could be stored or learned from.

2. No Voice Drift

  • AI must adapt to our voice and standards.
  • Do not allow tone, confidence, or authority to be overwritten.

3. No Final Authority

  • AI cannot approve facts, prices, dates, editions, or provenance.
  • These remain human responsibilities.

4. Reduce Friction, Not Thinking

  • Use AI to eliminate repetitive labour, not replace judgement, taste, or relationships.

5. Version Control

  • Save AI-assisted drafts clearly marked as such.
  • Final documents must be human-approved and archived correctly.

In Short

  • Perplexity → context & sourced answers
  • DeepSeek → discovery & deep file analysis
  • ChatGPT/Claude → polish, structure & creative ideation
  • Gemini (Gmail) → inbox speed

AI works behind us, not instead of us. And remember, as another titan of Chinese philosophy, Laozi (c. 6th century BCE), wrote in the Tao Te Ching: “A journey of a thousand miles begins with a single step.”

Read Michael’s first story on AI for Women Love Tech here. Artworks by Jill Daniels & Jane Reynolds. Credit: Michael Reid Art Gallery

Robyn Foyster: A multi award-winning journalist and editor and experienced executive, Robyn Foyster has successfully led multiple companies including her own media and tech businesses. She is the editor and owner of Women Love Tech, Women Love Health, and Women Love Travel plus The Carousel and Game Changers. A passionate advocate for diversity, with a strong track record of supporting and mentoring young women, Robyn is a 2025 Winner of the Samsung IT Journalism Awards. She is also a 2023 Women Leading Tech Champion of Change finalist, 2024 finalist for the Samsung Lizzies IT Awards and 2024 Small Business Awards finalist. A regular speaker on TV, radio and podcasts, Robyn spoke on two panels for SXSW Sydney in 2023 and Intel's 2024 Sales Conference in Vietnam and AI Summit in Australia. She has been a judge for the Telstra Business Awards for 8 years. Voted one of B&T's 30 Most Powerful Women In Media, Robyn was Publisher and Editor of Australia's three biggest flagship magazines - The Weekly, Woman's Day and New Idea and a Seven Network Executive.

This website uses cookies.

Read More