When AI replies too quickly, it removes the friction that makes thinking real. The result feels efficient, but something vital disappears: the space between question and understanding.
In this post we will:
• Notice how friction and delay create deeper thought.
• Practise a prompt that restores reflection.
• Reflect on how conversation changes when speed slows down.
Speed has become our default measure of intelligence. Yet insight rarely happens at speed. It happens in hesitation, in the second before the answer forms.
Last week, in Why AI Ethics Fails When Left to Experts, we explored shared responsibility. Some of your reflections appear at the end of this post.
This week, we turn to friction and ask what is lost when AI becomes too smooth to resist.
Step-by-step
After your AI tool gives an instant answer, pause and ask it:
Before replying, ask me one question that would deepen this thought.
Then:
• Notice what kind of question it asks.
• Write down your answer to that question before continuing.
• Observe how your thinking shifts when the tool slows down.
Speed looks like progress, but thinking lives in the pause.
If you are new to Slow AI, our first invitation was Teach It Something It Cannot Know.
A moment from me
When I was drafting a Substack not about AI and ethics, I asked ChatGPT:
What is the biggest issue people face with AI ethics?
Before hitting send, I added the prompt:
Before replying, ask me one question that would deepen this thought.
It paused, then asked:
Do you mean policymakers, developers, or individual users?
That question caught me off guard. I had been writing as if there were one audience, one shared ‘public.’ I should have known better from my own research in science communication. There is no single public, only many publics, each with different values, pressures, and kinds of responsibility. The same is true for AI and ethics.
Since then, I have built this pause into my process: one question before every answer. It slows the pace, but widens the view. Each exchange feels less like extraction and more like conversation.
If you have a moment, I would love to hear how Slow AI fits into your days. What draws you back, what you skip, and what you wish existed here.
What to do with it
If you want to share:
• Try the same prompt in your next chat with your AI tool.
• Notice what changes when you invite a question before an answer.
• Write down one small way you can build that pause into your routine.
If you share your reflection in the comments, tell us what happened when you slowed your thinking down.
Friction is not failure. It is focus.
If you want to keep learning slowly, and with purpose, try Cozora, a live community where we build thoughtful and useful AI together each week.
Why this matters
AI’s greatest temptation is speed. It makes us believe that fast answers equal deep understanding.
Yet the practice of slowing down to ask better questions is ancient wisdom. Long before search bars and chatbots, ancient Greek philosopher Socrates wandered the streets of Athens asking people what they truly meant by words like justice or truth. His method was simple conversation built on curiosity and friction. He believed that the hardest questions could not be answered quickly because understanding takes time.
The same lesson applies today. Insight grows in the pause, in the moment when confusion is allowed to breathe. When we slow a conversation with AI, we give ourselves the chance to notice what matters, what feels uncertain, and what might deserve another question.
The goal is not to make everything slower, but to remember that reflection has its own rhythm, one that every generation of thinkers has learned to trust.
If you would like to keep practising, you might like A Gentle Guide to Reflection with AI (pay what-you-wish). It offers simple exercises to help you pause, listen, and think with more intention.
From last week: Why AI Ethics Fails When Left to Experts
(thank you to all who shared)
John Brewton captured the heart of the discussion in one line: “Ethics breaks when it becomes performance instead of practice.” This reminds us that genuine ethics is lived through small, daily decisions, not drafted in policy documents.
Data Frank shared a practical insight from his own creative work, describing how adding a reflection step to his process by asking “Who is this really helping?” brought humanity back to his writing and reconnected him with his readers.
Sherry Heyl noted that ethics “isn’t a checklist or a department,” but “a shared conversation about how we live with technology.” This comment echoed a wider call for co-created responsibility rather than compliance.
Together, these reflections show that ethics thrives when it moves from paperwork to practice, from compliance to conversation.
If you try this week’s prompt, I would love to know what question your AI asked, and what changed when you answered it before moving on.
I read and respond to all your comments.
See you next Tuesday.
Go slow.
“Satnav brain” is the right metaphor.
The product design implication: keep a forced attempt before the model speaks. Same as desirable difficulty, just applied to AI UX.
I build adaptive reasoning practice around that sequence (hard distinction first; feedback after). Founder: https://intelligencemax.ai/guide
I thought it was just me feeling overwhelmed by AI responses. Even after I asked it to break it down for me the to do list was too overwhelming. Appreciate the practical approach to slowing down.