Reading Time: 8 minutes

Hey Prompt Lover,

Yesterday I asked one question and then waited.

If AI writes it and you send it, is it still you?

I didn't expect the volume of replies that came in. I genuinely didn't. And I want to share what you said before we move to today's question because the community's answer was more interesting than mine.

Luanne replied first. She told me the last thing she sent that was AI-assisted was an email to a roofing company she had negotiated with but decided not to hire. She went with a competitor and had to tell them.

She said something that stuck with me.

"If the person who received it knew, I think they would not have felt as positive about the message. AI helped me be more professional and kind with advising of my decision."

Read that again. She used AI to be kinder than she might have been naturally. The feeling behind the message was hers. The professionalism was AI. And she thinks the receiver would have felt differently knowing that. She is probably right. And I am not sure that means she did something wrong.

Tom Dunham is a PhD and the CEO of Bluegrass Consulting Services. He replied with something I had not considered.

He told me that sometimes when he uses AI he includes the prompt alongside the output. He sends both. He lets the person know his thoughts were sincere and that AI helped him communicate them better. When he uses AI to write a poem for someone he says so in the note. "This is a poem about you with a little help from AI."

Tom found a way out of the discomfort that I had not thought of. Transparency. Not hiding the tool but naming it. Not "is this still me" but "here is what I actually think and here is what helped me say it." That reframes the question entirely.

Then there was the reader who signed off as theazaland. They said something that genuinely made me stop.

"I dislike the confidence Claude has. Many times when it tries to correct my writing, especially those that convey real human empathy, Claude sees it as unnecessarily emotional."

And then: "But I stand my ground, teaching it that it's what makes us genuinely human. Then Claude cowers in rare humility and acceptance of its limitations."

"I teach it. I use it. I help Claude too."

That is the most interesting relationship with an AI tool I have heard anyone describe. Not using it passively. Not accepting what it produces. Actively pushing back when it strips the human out of the work and insisting on keeping it there. That person is not losing their voice to the tool. They are defending it inside the conversation.

Before I go into my verdict this newsletter is sponsored by Mintlify, check them out.

Last month, 48% of visitors to documentation sites across Mintlify were AI agents, not humans.

Claude Code, Cursor, and other coding agents are becoming the actual customers reading your docs. And they read everything.

This changes what good documentation means. Humans skim and forgive gaps. Agents methodically check every endpoint, read every guide, and compare you against alternatives with zero fatigue.

Your docs aren't just helping users anymore. They're your product's first interview with the machines deciding whether to recommend you.

That means: clear schema markup so agents can parse your content, real benchmarks instead of marketing fluff, open endpoints agents can actually test, and honest comparisons that emphasize strengths without hype.

Mintlify powers documentation for over 20,000 companies, reaching 100M+ people every year. We just raised a $45M Series B led by @a16z and @SalesforceVC to build the knowledge layer for the agent era.

Make Your Docs Agent-Ready

Based on everything that came back, here is what I think the community actually believes even if they didn't say it exactly this way.

The writing is yours if the thinking behind it is yours. If the idea, the care, the intention, and the judgment are genuinely coming from you — and AI is just closing the gap between what you feel and what you can say — then the communication is real. The tool is irrelevant.

The writing is not fully yours if you are using AI to generate the caring itself. To simulate a thoughtfulness you don't actually have. To produce a version of yourself that is more professional or more kind or more articulate than you actually are in that moment, not because you had the feeling and couldn't express it, but because you didn't have the feeling at all.

Tom's approach might be the most honest one. Name the tool. Let the person know your thoughts are sincere and AI helped you land them properly. Remove the question entirely.

The discomfort most people feel about this is not about authenticity. It is about disclosure. And disclosure is a choice anyone can make.

That is where yesterday's question actually lands.

What skill did you quietly stop building the moment AI could do it for you?

This one is harder than yesterday's.

Yesterday's question was about what you send. Today's is about what you are becoming.

Here is what I mean.

Every time you use AI to do something you used to do yourself, there is a version of that skill that doesn't get practiced. The writing muscle that doesn't get used. The research habit that doesn't get built. The ability to sit with a hard problem long enough to actually understand it rather than just get an answer to it.

Most people are not aware this is happening. It doesn't feel like losing something. It feels like gaining time.

But there is a specific type of cognitive atrophy that comes with outsourcing thinking. Not immediately obvious. Not painful in the moment. The kind you only notice when you try to do the thing without the tool and something that used to feel natural now feels unexpectedly hard.

I noticed it in my own writing. There are certain kinds of structural thinking I used to do naturally when building a long argument. I would hold multiple threads in my head and figure out the right order. Now when I try to do that without Claude my first instinct is to open Claude.

That instinct is the thing worth examining.

The skill I have most quietly stopped building is the ability to write a first draft I am not embarrassed by.

Before I used AI regularly, I could sit down and produce a usable first draft on most topics I knew well. It was messy. It needed editing. But it was structurally sound because I had developed the muscle for it over years of doing it.

Now my first drafts are worse than they used to be because I have stopped trusting them the moment I write them. The instinct to send it to Claude before I decide if it's any good has replaced the instinct to sit with it and decide for myself.

I got better output. I got a weaker draft muscle. I did not notice the trade until I was already in it.

That is what skill atrophy from AI assistance looks like in practice. Not dramatic. Not sudden. Just a slow drift toward dependence on a tool for something you used to be able to do without it.

Attio is the AI CRM that keeps you ten steps ahead.

Ask Attio anything. Where should I focus? What deals are at risk? Search, update, and create across your customer data.

Ask more from CRM. Ask Attio.

Try Attio for free

Think about the work you do every week. The writing, the research, the analysis, the communication, the problem-solving.

Now think about what AI handles for you regularly.

Pick one of those things. The one you used to be able to do well on your own and haven't done on your own in a while.

When did you last do it without AI? What happened?

Reply and tell me. And if you are willing to name the skill you think you are losing, I genuinely want to know. The most honest replies will shape Day 3.

Day 3 arrives tomorrow. The question follows from everything that has come up in the first two days.

— Prompt Guy

Day 2 of 30. One question every day that the AI industry is not asking. Forward this to someone who should be in this conversation.