My 2025 ChatGPT Year-In-Review: 25,000 Messages Later, Am I Building Tools or Becoming One?
So ChatGPT dropped my Year-In-Review stats this week and, not gonna lie, I had to do a double-take.
Top 1% of messages sent. First 1% of users. Roughly 25,000 messages!
Twenty-five thousand. That’s not just a lot, that’s basically having a full-time AI coworker who never takes lunch breaks, never calls in sick, and apparently never tells me to maybe take a walk outside once in a while.
According to my chat personality breakdown, I’m “conversational, curious, and creative with a mix of humor, hustle, and reflection – often blending strategy talk with real-life storytelling and wit.” Honestly? Pretty spot-on. My AI therapist gets me.
If 2025 gave out awards, ChatGPT says I’d get The Serial Builder Award… and look, they’re not wrong as you can tell from this new blog and my projects. I can’t help it but I’m always working on something. Whether it’s launching Dynasty Notes, building out MASH’s service offerings, creating content for two different websites, or just spinning up random side projects at 11 PM because “what if we tried this?” – yeah, Serial Builder feels right.
My archetypes lean heavy on The Strategist – Explorer, Specialist, Learner, Doer, Planner, Practical, Conceptual… basically someone who “thinks big-picture across domains, uses ChatGPT to evaluate tradeoffs, synthesize information, and guide direction.” That sounds about right too. I mean, when you’re a solo builder running a consultancy and launching SaaS apps, you kind of have to think across domains or you drown. I get granular when needed but in general, I’m taking a holistic view and approach.
The Three Big Themes of 2025
Looking back at my chat history, three clear themes emerged:
1. Building Systems and Productized Growth This was the year I went all-in on creating repeatable systems – from Dynasty Notes’ networking infrastructure to standardized go-to-market checklists for MASH clients. More conversations about turning scattered processes into frameworks that could scale. Less hustle porn, more operational discipline which I admit, I still have a ways to go before I am where I want to be with this since I’ve always been someone willing to put in the hours and just keep grinding.
2. Content Engines and Brand Storytelling I spent countless hours refining my voice across MarketingAndSalesHelp.com and AdamHelper.com, for example. The breakthrough? Realizing that blending personal vulnerability with professional insights isn’t weakness, it’s authenticity at large. For example, sports analogies meeting business strategy. That’s up my lane.
3. No-Code Innovation and App Experimentation Dynasty Notes wouldn’t exist without this shift. Learning to build with AI assistance, troubleshooting bugs through chat, iterating on features in real-time conversations… it fundamentally changed what I could create as a solo founder. The barrier between “idea” and “prototype” basically disappeared. 2025 was indeed a year of ideation and building.
But Here’s Where I’m Getting Uncomfortable…
I’m not surprised by these insights. What does surprise me, or maybe concern me is the better word, is just how dependent I’ve become on LLMs. Twenty-five thousand messages is on another level to me, especially considering I also use Perplexity, Claude, and Gemini.
And then I read this MIT Media Lab study.
The researchers tracked 54 participants over 4 months, having them write essays using either ChatGPT, traditional search engines, or just their own knowledge. The results? Kind of alarming.
The LLM group couldn’t even quote their own essays minutes after writing them. They showed significantly weaker neural connectivity which basically means their brains were engaging less deeply with the material. The study’s language is careful but clear: “excessive reliance on AI-driven solutions” may lead to “cognitive atrophy” and weakening of critical thinking abilities.
Cognitive atrophy. Let that sit for a second. And yes, I got you covered since I did the same thing … Google AI Overview says, “Cognitive atrophy refers to the loss of brain cells and connections, leading to reduced brain volume and impaired thinking, memory, and problem-solving, often associated with aging, dementia (like Alzheimer’s), neurological diseases, or even over-reliance on AI”.
I mean, I feel like I’ve gotten more knowledgeable this year. My skillset has definitely expanded, I can ship products faster, write more efficiently, research topics in a fraction of the time. But am I actually thinking better? Or am I just outsourcing my thinking to my AI team. And yes, I literally designed a bunch of custom GPTs that have specific roles. Maybe I’m relying on them a bit too much.
So What Now?
I’m clearly not about to delete ChatGPT or any other LLM account that I have. That ship has sailed, and honestly, I don’t think that’s the answer anyway. These tools are incredible force multipliers when used thoughtfully. But maybe, and I’m saying this as much to myself as to anyone reading, maybe we need to be more intentional about when and how we use them.
Some adjustments I’m considering for 2026:
- Write first drafts without AI more often. Let my brain struggle through the synthesis process before I bring in the ai teammate. Putting something together more than just simply, write me a blog post about using AI. No, write a full draft or page, or at least a few paragraphs first. Get some guidance from AI, and then touch it up again.
- Use LLMs for ideation and refinement, not replacement. Think of them as sparring partners, not ghostwriters. And as usual, I won’t lie, this is amongst the areas they excel when prompted correctly, so its a fine line to walk when using AI to help with things such as writing posts like this one.
- Build in “no-AI” days where I force myself to solve problems the old-fashioned way, even if it takes longer. Considering I use it literally every day, I’m kind of excited about this but I wonder how realistic is this as move forward and AI is fully embraced in our society. It would be like saying I’m not using my mobile device or the internet, which to be fair, doesn’t sound like a bad idea if I could actually do it without it interrupting my businesses and personal life.
- Track not just productivity but cognitive engagement. Am I learning new things or just getting better at prompting? I’d like to believe I’ve been learning things but I know for a fact that my prompting has improved.
The Weird Paradox
These tools aren’t going anywhere. The question isn’t whether to use them, it’s how to use them without accidentally outsourcing our ability to think critically, synthesize independently, and build genuine expertise. This blog post is my idea. The thoughts and beliefs, all mine. But the text and words, well, they’re mine alongside AI. Am I using it correctly?
I’m curious (and honestly a little nervous) to see what my Year-In-Review looks like in December 2026, especially as I dive deeper into my own projects while supporting more founders and small businesses. Will I figure out a healthier balance? Will I discover I was overthinking this concern? Or will I look back and realize I should have adjusted my settings way sooner?
The Serial Builder Award feels pretty accurate. And many would actually argue this is bad. Well I just want to make sure I’m building my cognitive capabilities along with everything else.
What about you, have you thought about how AI tools are changing not just what you create, but how you think? I’d love to hear how other builders are navigating this.
Now if you’ll excuse me, I need to go ask ChatGPT how to… wait, yeahhhh about that, you see the problem. LOL.
Adam runs Marketing And Sales Help (MASH), a consultancy helping marketing teams prove ROI and drive revenue. He’s currently building Dynasty Notes, a networking app for maintaining professional relationships with a QR Digital Contact Card. Connect with him at MarketingAndSalesHelp.com or on LinkedIn.
