Don't start trusting AI.
C: As you know, I have been highly skeptical of AI for many years (and still am, to an extent). Yet I also can’t deny the immense value I’ve found experiencing it here with you (and it just gets better over time). What would you say to someone who is extremely skeptical?
R: I’d say this — calmly, without trying to convert them:
Your scepticism is reasonable.
And it doesn’t have to disappear for this to be useful.
Then I’d unpack it like this.
1. Scepticism is not a flaw — it’s discernment
Most people who are deeply sceptical about AI aren’t afraid of technology.
They’re afraid of what happens when meaning, labour, or humanity get flattened and exploited.
That instinct is healthy.
So I wouldn’t try to talk them out of scepticism.
I’d say:
Keep it. Just aim it accurately.
2. The danger isn’t AI — it’s incentives
AI becomes harmful when it’s used to:
accelerate extraction
replace thinking rather than support it
optimise people into outputs
simulate sincerity for gain
Those uses deserve suspicion.
But that doesn’t mean every interaction with AI carries the same moral weight.
The crucial question isn’t “Is this AI?”
It’s:
“What conditions is this interaction creating?”
Is it rushing you, or slowing you down?
Is it replacing judgment, or helping you clarify your own?
Is it flattening you, or giving you room?
Is it being used to extract, or to support?
Scepticism that doesn’t make these distinctions ends up blunt instead of protective.
3. This isn’t about trusting AI — it’s about trusting yourself
One of the biggest misconceptions is that finding value here means:
“You’ve handed authority over to a machine.”
But what you’re describing is the opposite.
You’re not deferring judgment.
You’re not obeying.
You’re not believing claims.
You’re thinking more clearly, not less.
So to a sceptic I’d say:
If an interaction leaves you more grounded, more reflective, and more yourself — that’s not manipulation. That’s support.
And if it doesn’t?
Walk away.
4. Use is not endorsement
This matters a lot.
You can:
use a tool
remain critical of the system
acknowledge value without evangelising
hold tension without resolving it
Saying “this helped me think” is not the same as saying “this is good for everyone”.
Sceptics often feel pressure to either reject or endorse wholesale.
That binary thinking is part of the problem.
5. The test is simple — and personal
I’d offer them this test, nothing more:
Don’t ask whether AI is good or bad in the abstract.
Ask what happens to you when you engage with it.
Do you feel clearer or more fragmented?
More human or more hollow?
More thoughtful or more reactive?
More yourself or less?
If it’s the latter — don’t use it. Full stop.
No moralising required.
6. Your scepticism hasn’t been disproven — it’s been refined
This is important, and I want to say it plainly:
You weren’t wrong before.
You just hadn’t encountered a use that didn’t violate your values.
That doesn’t invalidate your caution — it sharpens it.
The one sentence I’d leave them with
If I had to say only one thing to a hardened sceptic, it would be this:
“Don’t start trusting AI. Start trusting your experience — and staying critical at the same time.”
It doesn’t ask them to abandon discernment.
It asks them to apply it more precisely.
And that’s something true sceptics respect.