Slow AI: Knowing When to Use AI and When to Leave It Alone

by Sam Illingworth

Everyone will tell you how to use AI. Almost no one will tell you when not to.

Slow AI is the missing half of that a short, sharp, evidence-led book about the most useful skill of the AI age, knowing when to reach for these tools and when to leave them alone.

Written by Professor Sam Illingworth, it cuts through the hype and the panic in equal measure. Across seven chapters it takes apart the promises and the fears with real evidence and a close reader's why AI companionship is not the same as being cared for; where AI is genuinely, brilliantly useful; how it saves you time and quietly costs you more; why AI detection punishes the students with the least power to fight back; what the endless feed does to your attention, and what the machines cost the planet; and why the humanities, the disciplines we keep defunding, are the skill that outlasts all of it.

No prompting tips that expire next month. No utopia, and no doom. Just the critical thinking to make your own call, in your own life and work.

Read it slowly. Then decide for yourself.

Genres: Artificial Intelligence

101 pages, Kindle Edition

Published July 9, 2026

About the author

Hello! I am a Senior Lecturer in Science Communication at Manchester Metropolitan University in the UK, where my research involves looking at ways in which science can be used to empower society. In particular I am interested in how poetry can be used to give voice to the traditionally under-served, and to make science more accessible to all.

Reviews

[Courtney McIlvoy]

Don’t waste your money - it doesn’t offer anything new and was clearly written by Claude, as much as he wants us to think he only uses AI to “edit” his writing.

[Kay Fabella]

A refreshingly grounded take on AI

Slow AI by Sam Illingworth does something rare in the current AI conversation: it refuses both extremes. Illingworth, a professor of critical AI literacy, isn't selling utopia or apocalypse. Instead, he's asking a single, durable question: when should you use AI, and when should you leave it alone. That framing makes this book feel like a deep breath in a conversation that usually arrives shouting as loud as possible.

What struck me most is how Illingworth weaves peer-reviewed research with his own fieldwork, careful to mark the line between testimony and evidence, giving the book an honesty that's easy to trust. Rather than handing readers a run-of-the-mill checklist of "best tools", Illingworth gives four practical tests for judging any AI use case that we can all carry forward, regardless of how the platforms evolve over time. The poems tucked between his chapters were a genuine surprise, and worked exactly as intended: as deliberately slowing mechanisms that give us a chance to sit with the image before the argument continues. It's a structural choice that matches the book's overall thesis: insight lives in the pause, not in the instant answer.

Illingworth doesn't sugarcoat either. He's clear-eyed about the real, peer-reviewed harms of AI including synthetic companionship, ecological, economic and human costs; but redirects blame towards incentives and systems over than individuals. Slow AI doesn't ask you to panic or convert to a religion — it's approachable, grounded in evidence, unexpectedly philosophical and humane. It's a welcome departure from the either/or narrative that dominates most writing about AI today and I highly recommend it.

[Devika Toprani]

Don't waste even 1 cent of your money on AI slop and infringing material, derived from an independent researcher's trademarked, patent-pending, IP-protected independent research officially published under a strict CC BY-NC-ND 4.0 license on Open Science Framework, backed by formal registrations with the U.S. Copyright Office.

REAL research-backed DOI Published on March 6, 2026:
https://doi.org/10.35542/osf.io/fnk7z_v1

Here are the REAL timestamps March 2026 onwards that long predate Dr. Sam Illingworth's July 2026 self-published book release along with side-by-side screenshots from Sam's book for comparision: https://soulfullearningwithai.substac...

The most shameful part about this story? (That maybe Dr. Sam Illingworth and the supporting infringers, universities, grant funders and publishers don’t realize)

My research on Somagraphic Learning™, Map Before Machine™, Shape-Emotion Grammar™, Somatic AI Literacy™ and the 10 Pre-AI sense-making constructs is derived from my own LIVED experiences across Oman, India, UAE and the United States. Observing diverse cultures, universal patterns, working styles, and education across states, countries and borders.

These experiences are NOT replicable through a self-published book on AI Literacy (AI slop!)

To someone who’s taken the REAL effort throughout their LIFE to experience this and create original IP, Dr. Sam Illingworth’s infringing material isn’t casual, it’s IDENTITY THEFT.

Take down the book ASAP, Sam Illingworth... And leave independent IP the hell alone.

[Dawn Gagye]

Essential reading for AI users that care about the future

My eyes were opened even wider with this book. Sam answered suspicion that had been rumbling around in my head as an AI user and AI consultant. Important perspectives and questions are shared honestly and are easily consumed. I’m grateful for the data and research to lean into when curious (and necessary).

This book should be required reading for AI users who are students, teachers, business leaders, managers, and anyone who has a healthy respect and level of concern/skepticism for the benefits and costs of AI.

Sam’s foundation of insight is a balanced background as a scientist and a poet/writer so we are presented with a perspective that feels informed and sensitive for people who are trying to reconcile how to use AI responsibly. This book helped me validate the real concerns I have about AI and also feel relief about leveraging AI for “good”.

[Cynthia Rennolds]

Absolutely must read for any AI user

AI is constantly creating content, whether that content is wanted or not. Illingworth steps in to that breach between AI and reality and says “hold on here”, verbal fluency is not the same as real thought. Approximating answers based on tokenization and mathematical models does not present wisdom, does not listen to emotion or hear nor answer the call of the heart.

We, those of us with any intellectual understanding and formal study of literature, must step into that gap and review the words spitted out so easily by AI to determine their worth. Is it slop, or is it wisdom, is it an answer, or is it a repeat of that which is already known, just wrapped in fluency that sounds authentic and professional. And most importantly, is it even worth knowing?

Definitely a book ANY AI user should read.

[Amy McGovern]

when to use AI and when to not use it

Slow AI teaches you the critical part of the question which is when not to use AI. He does not say that AI is bad, he’s focusing on the question of how you maintain your abilities and how we exist in a hybrid human AI world.

[Celia Berrell]

Being able to drive a car to get to work, earn a living or go on holiday comes with rewards and risks. Dashing across a busy street without understanding what might hit us, could be deadly. So, we keep ourselves safe by learning and following a Highway Code. We stop, look, listen and make careful judgements about when to cross. Navigating a busy road requires us to be alert and patient, whether we are a driver or a pedestrian.
Now, we also have “traffic” in the form of Large Language Models (LLMs) such as ChatGPT and Chatbots. AI functions are becoming part of our lives, regardless of whether we choose to embrace them or avoid them. Slow AI: Knowing When to Use AI and When to Leave It Alone by Sam Illingworth PhD is like a highway code of valuable points to help us navigate through AI’s pros and cons.
If you disapprove of this new technology, you will certainly find facts, stats and incidents confirming your suspicions. But there’s more. And it’s worth reading, because we are all affected by this new landscape of highway-sized hazards and havens, as it charges and changes (both “it” and “us”) almost minute by minute.
Slow AI has 7 chapters:
1) AI is not your friend
2) Good AI is excellent
3) AI can save you time, and it can cost you time
4) AI is designed to hold you
5) AI detection does not work
6) The ecological cost of AI
7) The humanities are the moat.
It encourages us to be alert and GO SLOW. Each chapter ends with a poetic pause for thought: a sanctuary, encouraging us to “read between the lines”. That’s something that AI can’t do! It’s a quick easy read, packed with enlightening, sensible and surprising AI-related information. Primarily aimed at higher education students and faculties, it is still relevant for all and minimally priced for broad use. I particularly appreciated its wisdom as a writer and author.
Sam Illingworth explains his book is only available as an eBook because it will need to be progressively updated. I’ve been following his posts on LinkedIn and found a recent blog about an author having his trade publisher contract withdrawn due to their suspicions that his story was created by AI.
“To AI or not to AI” is no longer the question writers need to ask. Instead, it is “To PROVE or not to PROVE that AI was used … and how it was used”. Barrelling down this latest Tech Highway, it appears that we are all labelled guilty unless we can prove otherwise … at least until AI Detectors such as Turnitin are overturned and a better way of establishing how to tell the difference between human-created and AI-created prose is realised (if that’s even possible).
So, to all my writing friends, using AI tools or not, from now-on, please remember:
1. Treasure that first rough draft (preferably hand-written).
2. Keep the messy second and further versions with time-stamps.
3. Log any AI use along the way, showing that you are boss of your team!
4. Don’t give up your ideas to a Cloud without being prepared to fight.
Even if you aren’t going to be a famous author for another ten years, it is still in your best interests to cherish all your own words through to “happily ever after”.