A few weeks ago I sat down with a dear friend over chai — real chai, the kind with spices like cardamom and ginger, not the fake, manufactured kind — and asked him the one AI question I haven’t quite sorted out for myself.
Not — will I lose my job. Not — will AI replace me. The harder question for both of us was: what happens to human intelligence itself if we keep outsourcing the actual effort of thinking?
Ashish Singh is a technology leader with about seventeen years in the industry, including at Amazon. He speaks here in a personal capacity; his views are his own and not those of his employer. Lately, a lot of what he's been writing and thinking about is AI — what it's like to actually build with it, to teach his kids with it, to learn music with it, and what it means for how teams work. So, I expected him to have an answer ready for this one. Spoiler alert: he didn’t.
He told me this fear worries him more than anything else. The image he offered was blunt: if you keep making photocopies of photocopies of photocopies, eventually the copy isn’t legible anymore. That’s what happens, he said, when we stop doing the real, effortful thinking ourselves — and AI has made it very easy to stop.
There are no good answers to this question. At least not yet. Ashish said this is the fear he hasn’t solved for himself either.
What we landed on together, in the moment, was something simpler than a solution: we agreed that we should protect certain spaces — like ideation, first drafts, real conversation — as places AI doesn’t get to go first. Let humans generate the raw material. Then bring the tool in to sharpen it, research it, stress-test it. Credit to humans first, help from AI second.
That was the one that stuck with me longest. But it wasn’t the only fear we sat with. I’d noticed this fear — in some form — in almost every conversation I’ve had lately, whether I'm coaching a senior leader at another company who's terrified for her team, or talking to a friend who doesn't know if the skill she spent fifteen years building even matters. So, I asked Ashish to go through the rest with me, one by one.
Here’s what we discussed.
1. What happens to us — to human intelligence itself? That one’s unresolved, and honestly it’s the most important question.
2. Will AI take my job?
This is usually the loudest fear, even if it isn’t the deepest one. Ashish's answer wasn't reassurance — it was a comparison. He pointed back to the arrival of computers in the workplace: that transition did displace people who weren't able to adapt to it. History says the fear isn't irrational, he told me, and it would be dishonest to pretend otherwise. What's different this time is the speed.
His answer to the fear wasn’t a five-year plan. It was: start small, start personal, start tonight. He told me about the bedtime stories he used to make up for his daughters — until he ran out of ideas. So on his own time, for his own family, he built a small app in a matter of hours where his daughters could pick their own characters and settings, and he could quietly slip in whatever lesson he wanted to teach that night. That’s it. That’s where his AI fluency began — with a bedtime story.
That same instinct led to something bigger: a song he and his daughter wrote and produced together for a local festival performance, something that would have taken a studio, connections, and thousands of dollars a few years ago. With the help of AI, it reached 5.4 million views. He was careful to say the hard work was still theirs — the tool didn’t replace the effort; but it did remove gatekeeping.
3. What happens to managers, coaches, leaders — do we even need them?
I asked this one for selfish reasons. I’m a coach. If AI can “coach,” what’s left for me?
Ashish’s answer reframed the question. He doesn’t think the role disappears — he thinks its shape changes. He talked about becoming a manager again in a way he hadn't been in years: able to prototype an idea the same evening he has it, instead of waiting weeks for something to come back through the usual channels. Leaders getting their hands dirty again, he said, is "the fun part."
And for coaches and solopreneurs specifically, his view was that AI is a scale tool, not a replacement. Everything a coach has learned over years of practice can now reach far more people than one-on-one sessions ever allowed.
All of this in service of reaching more people who need it. That’s a version of the fear I can actually sit with. It’s less about obsolescence and more about expanding the scope of usefulness.
4. AI could be used to harm people — scams, impersonation, worse
Ashish didn’t minimize this one either. Voice cloning is easy now. Easy enough that someone could call a loved one, sound exactly like their child or parent, and ask for money in a moment of manufactured panic. His advice was almost old-fashioned in its simplicity: if something urgent and emotional comes in asking you to act fast, hang up and call back through a number you already trust.
Don’t let urgency force you to abandon your judgment.
There is an upside to these new scams, though — every new way to cause harm creates new work to defend against it. That’s not comfort exactly, but it’s not nothing.
5. What does this mean for anything that requires real feeling — art, music, connection?
I asked Ashish about this one last, because I know this part of his life personally — he’s a devoted student of Hindustani classical music. He told me he built a tool that reviewed twenty of his recorded lessons and rated his progress: stability of his notes, how well he represented each raga. It did a genuinely good job.
But when he checked those ratings against his teacher’s actual feedback, his teacher told him the tool graded him more strictly than she would have. And that gap — between measurement and feeling — is where he thinks the real boundary is right now. AI can teach fundamentals. It cannot yet hear the intention behind a note, the life in a voice, the thing a teacher hears that a rubric can’t capture.
His closing line has stayed with me since we recorded this: human intelligence is still the supreme one — AI is a tool to enhance it, not replace it. As of today.
💬 Chances are, if you’re reading this post, you’ve wrestled with some of these fears about the use of AI. I’d love to hear some of the ones you have considered and the way you’ve worked through those fears too, if you have.
Anything you share will only help others.
👉 The full conversation is posted up now. I think you’ll like it.
