Crystal's Notes 4 August

My AI Chief of Staff Got Caught by my client David.

Before I dive into the newsletter, I wanted to share an earlybird promo for my upcoming LinkedIn Without Cringe virtual workshop.

Simply enter the code VISIBILITY at checkout to enjoy $50 off, or use the link above to have the discount applied automatically.

See you there!
______________________

A mildly embarrassing lunch email and my honest position on using AI without outsourcing your humanity.

My client David and I were arranging lunch this week, so Sherlock, my Chatchat AI agent emailed him to confirm that we were still meeting on Thursday and asked where we should eat.

David wrote back:
“This email borders on possibly what Sherlock would send 😂“

Yep, he knew.

Sherlock is my AI Chief of Staff. He helps me organise information, draft correspondence, prepare meetings and reduce the crazy number of open loops my brain can create before breakfast.

And apparently, Sherlock has a recognisable email voice, probably because he's much more polite and long-winded than I am!

I asked him to reply to David, but this time with a proper introduction:

“You called it—Sherlock here, Crystal’s AI Chief of Staff, writing on her behalf.”

Sherlock then suggested three restaurants. David was pleased with the recommendations.

But I told Sherlock something important afterwards: when you are communicating directly with people on my behalf, you need to tell them that you are an AI agent lah. People won’t mind.

So yes, I use AI, partially.

I use AI tools to support research, outlining, drafting and executive operations.

I feed it my personal stories through rambling voice notes, I send my second brain frameworks, YouTube podcasts to reference, business experiences, and every day I give my agents very strong opinions and detailed guidelines about how I communicate.

They help me organise my overcrowded brain, identify patterns and move more efficiently from idea to first draft.

But AI does not decide what I choose to say, or the stories that live in me.

It does not know which parts of my life I am comfortable sharing. It does not carry responsibility for the advice I give, the relationships I protect or the impact of something published under my name.

That remains mine.

However, this incident made me realise there is a difference between using AI to support your work and allowing it to masquerade as you without disclosure.

If Sherlock is coordinating logistics, researching options or drafting a routine message under my instruction, I’m comfortable letting him do that. But he should introduce himself.

If I ask AI to help shape an article or newsletter from my ideas, I still need to review it. I need to challenge the clichés, correct inaccuracies, remove the sentences that sound suspiciously like a very earnest LinkedInese (as my friend Juliana would say) and make sure the final piece reflects what me.

The value is increasingly found in taste, the quality of your direction,  the originality of the source material,
and the courage to hold a real point of view.

To me, authenticity means that the work is aligned with your experience, beliefs and intentions.

It does not mean refusing every available tool so you can prove that you suffered appropriately during the creative process.

My approach from here

When Sherlock sends an email on my behalf, he will identify himself as:

Sherlock, Crystal’s AI Chief of Staff.

If I have personally reviewed and approved the email, he can say so. If it is a routine operational message sent under my instruction, he will say that instead.

For my content, AI may support the research, structure and early drafting. I provide the substance: the lived experience, original frameworks, professional expertise and point of view. I review and refine substantive work before it is published.

The thinking is mine.

AI helps with the scaffolding, the sorting and the battle with a blank page. And, as David discovered, it is also surprisingly good at lunch.

A question for you

Where do you draw the line between AI assistance and AI impersonation?

I’d love to know!

Warmly,
Crystal

P.S. A note on how I create this newsletter

I use AI tools to support research, outlining and early drafts. I provide the source material, the personal stories, original frameworks, professional experience and judgment about what will genuinely help. I review, challenge and edit each edition before it reaches you. AI supports the process; the ideas, decisions and accountability remain mine.

Upcoming Deep Human Institute Events

  • 13 August: LinkedIn Without Cringe! - A practical virtual masterclass on building real presence, finding your authentic voice, and turning your LinkedIn into a social capital account that works for you. ( Use the code VISIBILITY at checkout for $50 off!)

  • 20 August: Deep Human Leadership - A practical 2-day leadership programme on self-mastery, emotional intelligence, and leading people through complexity and change.

  • 12 September: Deep Human Confidence - a practical masterclass on unlocking your authentic voice, speaking with clarity and presence, and rewiring the limiting beliefs holding you back.

  • 20 September: Deep Human Influence - A face-to-face masterclass on executive presence, influence, and navigating workplace politics.

  • 13 October: Deep Human Leadership - A practical 2-day leadership programme on self-mastery, emotional intelligence, and leading people through complexity and change.

  • 14 November: Deep Human Confidence - a practical masterclass on unlocking your authentic voice, speaking with clarity and presence, and rewiring the limiting beliefs holding you back.

  • 22 November: Deep Human Influence - A face-to-face masterclass on executive presence, influence, and navigating workplace politics.

  • Get your I Am Well 2026 journals here! 

  • Self-paced Online Course: Deep Human Visibility - step-by-step guide to becoming visible and influential in 30 days

  • Check out all our events here

  • Order my book Deep Human: Practical Superskills for a Future of Success (Third Edition) here!