76 Comments
User's avatar
Alex Pawlowski's avatar

I first read “ruin” and not “run” 😂😅

Expand full comment
Wyndo's avatar

I would have to shut down my substack if it was true 🤣

Expand full comment
Alex Pawlowski's avatar

I would love to team up with you and treat it like a thought experiment "how many AI Agents does it need to ruin your life?" hahaha...

Expand full comment
Wyndo's avatar

Haha I think u only need 1: an AI agent using ChatGPT memory which have access on your LinkedIn and post on your behalf and reveal your darkest side LOL

Expand full comment
Alex Pawlowski's avatar

Sounds goot to me…

Expand full comment
Betty Lee's avatar

SAME

Expand full comment
Steven Adler's avatar

Same! And I was excited to read that article

Expand full comment
Joseph de Castelnau's avatar

Isn't that the same thing for him?

Expand full comment
Daria Cupareanu's avatar

Really like how you framed this, especially the shift from “use AI more” to making it the default response to friction. A lot of the habit stacks you mentioned are similar to what I’ve started building as well. Once you integrate it into your workflows like this, it stops being a separate tool and just becomes part of how you operate.

Expand full comment
Wyndo's avatar

Indeed.

The major shift happens when you stop thinking “how i should be using ai more,” but it just happens naturally without noticing it.

Expand full comment
Michael Colligan's avatar

What AI agent would you suggest for scraping subreddits (and other data)

Expand full comment
Wyndo's avatar

I use Gumloop, but there are other tools like Make and n8n that most people also use.

Expand full comment
Ishwar Jha's avatar

It seems as if you are running and switching between AI tools more often. I am not sure if this can work for everyone, but good for you.

Expand full comment
Wyndo's avatar

Yes since we are not yet arrive on one tool for everything. But, 70% of this workflow happens on Claude. So, it’s still fine with me with generally low tools switching.

Expand full comment
Jennifer Hong's avatar

I'm an ardent fan of Atomic Habits. And since I'm fascinated by AI, this post hits all the marks at once. Have a similar sorta post upcoming, about how I use AI to write with me and not for me. Anyway, it's great to see this kind of thinking has a home (and audience) here.

My fave part of the gym analogy was the story of the man who showed up at the gym daily, didn't work out, and kept doing that until he did start working out. Showing up consistently was the first hurdle to jump.

Expand full comment
Wyndo's avatar

Thanks Jennifer, love the analogy.

Start from small, iterate, and soon enough it becomes habit and part of who you are.

Expand full comment
DCWASHING's avatar

Wonderful. Especially how you organized example a d why the method work. Just reading this got me past the one frustration. I am saving this and doing each step per week minimally but also when I need it. I need all 15. Thanks.

Expand full comment
Wyndo's avatar

look forward to see what are u going to implement soon :)

Expand full comment
DCWASHING's avatar

Fyi. Will be moving to a new alias account but will re subscribe

Expand full comment
DCWASHING's avatar

Great book on habits, too.

Expand full comment
Raymond's avatar

This is insightful. Thank you, Wyndo.

Expand full comment
Wyndo's avatar

Appreciate it :)

Expand full comment
Greg Wolford's avatar

Enjoyed this. I’ve reached the point where AI is my go to for a lot of my work. My next step is automation. I’m looking for ways to automate my workflows, especially research and writing. This takes some upfront effort but should pay off in the long run.

Expand full comment
Wyndo's avatar

yeah automation and interconnection is the next shift and i guess this is why AI agent takes world by storm these days.

have u checked my last post about MCP? that could be starting point. curious what are things you'd like to automate?

Expand full comment
Greg Wolford's avatar

I will check it out. Initially, I want to automate daily research and subscriptions to pick and choose which to go deeper on and comment on.

Expand full comment
Dr Jen | Syringa Wellness's avatar

Perhaps this is my seasoned perspective talking, but sometimes I wonder if all these cognitive extensions are just a sophisticated way to avoid sitting with our own thoughts. Sometimes it’s okay to just be a human without a dashboard.

Expand full comment
Subramani Aatreya's avatar

Hi Wyndo,

Your post initially overwhelmed me! I found myself immediately researching APIs and ways to integrate AI tools into everything from emails to calendars. However, after reading the latter part of your article, I realized the importance of starting with "low-hanging fruit." I've begun by integrating AI for email summaries, focusing on cost-effectiveness since I'm not yet monetizing my content creation.

This post has truly been a wake-up call. I now understand that AI tools are not just motivators; they are powerful habit stackers. Thank you for this insightful realization!

Expand full comment
Wyndo's avatar

Yes!

I think you totally nailed this :)

You dont have to overcomplicate things and start from easiest thing you can do right away and move up accordingly.

Glad you resonated with this!

Expand full comment
Thu Anh Nguyen's avatar

Thank you for compiling this advice! All the common mistakes you introduced resonated with me and reminded me of why it’s been so hard to integrate the models into my life. I’ve been trying to become a better writer and communicator and this definitely gives me a better understanding of what I can do to introduce the models seamlessly into my life.

Expand full comment
Wyndo's avatar

glad you find it useful :)

I used to overthink how I could integrate these models seamlessly on my workflow and life, but it just created analysis paralysis and ended up I didn't so much things.

But, by starting from low friction work, I can integrate it slowly and moving up as I have more understanding what I want. Massive game-changer!

Expand full comment
Roi Ezra's avatar

This is spot-on. AI isn’t just a shiny tool we occasionally pick up, it’s a profound shift in how we structure our thinking and approach our daily challenges. In my experience, effective human-AI integration isn’t about using more tools, it’s about deeply embedding reflection, clarity, and intentional systems into our workflow. As I’ve explored, what we’re really experiencing isn’t replacement by AI, it’s a redefinition of meaning in our work and lives. Your Atomic Habits approach brilliantly captures how powerful it can be when we build AI into intentional routines rather than chasing novelty. Thank you for articulating this essential perspective so clearly.

Expand full comment
Wyndo's avatar

Much appreciated, thanks Roi :)

Expand full comment
Heather Hoskison's avatar

Anytime i need to Google something... I dont. So far so good. 👍🏼

Expand full comment
Paolo Perrone's avatar

"Since Claude already has access to my newsletter archive.." how do you do this?

Expand full comment
Wyndo's avatar

First, upload newsletter archive on GDocs and connect it through project knowledge so it can automatically sync. So Claude can read it and make connection with new data you feed them into.

Expand full comment
Paolo Perrone's avatar

thanks! really appreciate it!

Expand full comment
Fig's avatar

I think AI will soon will allow us to build software even with just prompts in English

Expand full comment