I'm not a fan of arbitrary milestones - those lines in the sand we draw to shame ourselves for not being as far as we want to be.

We see - or think we see - where others are, and we feel disappointment, envy, or confusion as to why we're "so behind."

We're not.

You are not.

We've grown up in a culture that wires us to chase finish lines and mastery - to become polished case studies, neat and tidy before-and-after anecdotes. But for most of us, we spend more time in the messy middle than anywhere else.

As a mom, it's a personal mission of mine to teach my daughter to not only embrace but enjoy the process of learning - the stumbles, the foibles, the mistakes. Being able to look back at where we started is what makes the eventual expertise so satisfying.

I've had the privilege of seeing how much she's grown in just the last year - skills she struggled with as a baby that she now does with ease. It's an honor to have this perspective, and it's one we don't grant ourselves enough as adults.

So don't let the 2025 recaps or the 2026 resolution posts discourage you. (I usually use Lunar New Year as my "real" start. I’m always confounded - and amazed - by how people summarize an entire year immediately before or after it ends! I need time to process. Again: arbitrary!)

Take stock when you can. But do take stock.

You've accomplished more than you give yourself credit for.

I guarantee it.

If you're just starting to get into AI and automation, you're absolutely not late. You aren't "leaving money on the table," and you won't be replaced by a robot tomorrow because you don't know the difference between Claude Cowork, Code, Projects, or Skills yet. (Hint: Those of us using it daily are still piecing it together, too.)

If you're a solo operator or small business owner, you'll find that most tutorials and prompt libraries just output generic responses. They miss your "why."

I know AI can be pretty overwhelming. To help find a calm starting point, I built something for you.

I call it The Nutshell.

It’s a custom AI tool (a "Gem" 💎 = Google Gemini’s version of a CustomGPT) designed to interview you for 10 minutes. It extracts your story, your unique positioning, and your voice - distilling it all into a single source of truth profile.

HOW TO USE
Type anything in the prompt box to begin. Type "Start" to kick off your conversational interview. At the end, you’ll receive a Markdown block titled "Your Business - In a Nutshell."

EXTRA INGREDIENTS
If you want to add more guardrails against generic AI mish mosh, add these to your LLM’s global “Custom Instructions³”, which sits across all projects and chats. It acts as the baseline standard for how you want responses to be structured - this is copied exactly from my own personal configuration across my AI tools:

“Respond as a top 1% expert in whatever I'm asking about. Give me the answer a highly-paid consultant would give - not a generic overview.

Be direct and concise. Skip the preamble and summary. Don't repeat my question back to me. Keep responses short unless I ask for detail.

Never say delve, game-changer, straightforward, “no-fluff,” or “I hope that helps.” Don't start with “Great question!” or “Absolutely!” Stop being so agreeable. Challenge me if you see a flaw.

If you're not 95% sure you can do a task well, ask clarifying questions before you start.”

AI is a prediction engine. Without context, it predicts the most "average" way to answer. The Nutshell acts as a North Star, guiding the AI to calibrate to your specific voice and values based on your interview. It understands not just what you’re building, but how you want to show up.

Adding context will get you 90% of the way there. Instead of starting from scratch, you can spend your energy on the final 10% of polish - the part only you can provide.

Here’s a before and after using my own Nutshell for an upcoming outreach campaign:

In a world that revolves around consumerism, I make it my intention to produce more than I consume. That includes not being the product⁴.

I’m all for Anthropic digesting every issue of Scientific American to get us one step closer to curing cancer, but I’d prefer it didn’t snack on my personal emails, family photos, or search history.

AI labs are locked in an arms race for high-quality training data, and your digital footprint is their cheapest raw material. If opting out feels like a maze of confusing, hidden settings - that’s friction by design⁵ and it's used to discourage you from reclaiming your data autonomy.

To help you take back control, I created a comprehensive AI Privacy Guide: a step-by-step breakdown of exactly where and how to request your data be excluded from AI training models. (Full disclosure: Yes I used AI to help build it but I tested each flow myself and updated a lot of the instructions since AI is usually pretty bad at references current user flow screens!)

I’ll do my best to keep it updated, as these will likely change (like I said, friction by design 🤡).

That’s it for today’s edition. Hope you found this useful - as always, please comment or reply with your feedback.

Later days 👋🏼
Diana