How I Avoid Being an Overall Terrible Person Using a Weekly Accountability System That Doesn’t Suck
Hint: IT INVOLVES AI. I KNOW, BEAT A DEAD HORSE.
Left to my own devices, I will spiral.
I'm talking full-blown "Is it Tuesday or March?" mode, surrounded by empty ramen bowls, unread emails piling up, and at least one major life decision being avoided via just one more round of Street Fighter 6 — which I’ve somehow convinced myself is research for a newsletter piece about, I don’t know, success? Or discipline? Something vaguely redemptive?
It's not pretty. It's not even efficient. But it's very, very me.
That's why I built a weekly accountability system — not to maximize productivity or crush goals or whatever LinkedIn influencers think I should be doing — but to avoid becoming a complete car crash of a human being.
You know, the kind who forgets birthdays, ignores texts for weeks, and turns life into an endless blur of anxiety-ridden days that all feel exactly the same.
The twist? This system doesn't make me want to throw my laptop into the ocean.
Wait, what's wrong with normal accountability systems?
Nothing, I guess. Okay, everything.
Most of them feel like they were designed by people who I’m sure are absolutely delightful. But they’re people who have never spent four hours doom-scrolling or eaten an entire bag of chips while staring blankly at a Google Doc.
There’s too much rigidity. Too many checkboxes. Too much earnestness.
I needed something messier — something that acknowledged I'm a walking contradiction. Capable of hyper-focusing on the most random things while simultaneously forgetting to drink water all day. Something that could handle the truth of me: a human doing their best while occasionally self-medicating with distractions (some legal, some… let’s just say state-dependent) and still trying to grow.
The Slightly Snarky Consultant Prompt™: My Secret Weapon
Here's how it works: I journal almost every day. Sometimes, it's profound and insightful. Sometimes, it's literally:
ordered takeout again.
why can't I cook like a normal adult?
jfc i'm in my late 40s and i watched three episodes of a show I don't even like.
what am I doing with my life oh goddddd
At the end of each week, I feed my entries into an AI prompt I call the Slightly Snarky Strategy Consultant.
This entity is part therapist, part management consultant, and part that one friend who hugs you with one arm and slaps your shoulder with the other, saying, “Get it together, sweetie.”
I do this using Obsidian and a plugin that lets me run AI like Claude or ChatGPT on multiple daily entries embedded as notes — I use Caret Chat, but other plugins do something similar — and I feed it a query that looks something like this:
[[Slightly Snarky Strategy Consultant Prompt.md]]
- [[2025-03-24]]
- [[2025-03-25]]
- [[2025-03-26]]
- [[2025-03-27]]
- [[2025-03-28]]
- [[2025-03-29]]
- [[2025-03-30]]
- [[2025-03-31]]If you know your Obsidian, you know that each of those dates inside the double square brackets are actually markdown files filled with AI prompts or, you know, salacious diary entries. Pick your poison.
And of course, you don’t have to use an LLM plugin with Obsidian. You can use a general AI chat site and paste in journal entries directly — just expect a lot more copy/pasting.
Here's what it gives me back:
🔍 Connect the Dots
This is where the AI lovingly drags me.
It highlights patterns I didn't notice (but absolutely should have). Like:
“You've mentioned being ‘exhausted but still on Twitter at 2am’ four times this week.”
“There's a recurring theme of avoiding your mom's calls while simultaneously worrying that something bad will happen to her.”
“You keep saying you’re fine, but you've used the phrase 'internally screaming' six times since Monday. Interesting.”
It doesn't just tell me what I did — it tells me the story I'm telling myself.
💡 The "Oh, That's Why" Moment
This part always hits too close to home.
It digs underneath the surface behavior and finds the messy human stuff I'm trying to avoid.
Truth bombs such as:
“You're not procrastinating because you're lazy. You're procrastinating because you're terrified of not being perfect.”
“You're avoiding your family not because you don't care but because caring feels like drowning sometimes.”
It's never cruel. But it is... annoyingly accurate.
🪜 From Insights to Baby Steps
For every insight, I get one tiny action I can take.
Something so small it short-circuits my overthinking. Think:
“Text your mom a single emoji. That's it. One emoji.”
“Put shoes on and walk outside for exactly three minutes.”
“Write down the thing you're avoiding on a sticky note and stick it somewhere you'll see it. Just acknowledge it exists.”
These steps are always slightly ridiculous, and that's the point. If I laugh, I'm more likely to follow through.
🤘 Next Week's Tiny Rebellions
This section asks: “How can you gently tell your patterns to go fuck themselves?”
I get 2–3 micro-acts of rebellion, each under 15 minutes long and designed to disrupt my well-worn paths of avoidance. Examples:
“Delete Instagram for 24 hours. Just 24. Not forever.”
“Eat breakfast sitting down, not hunched over your keyboard.”
“Call someone instead of texting. Experience the horror of real-time human connection.”
Sometimes, they're dumb. But present-me loves future-me for doing them.
🙌 The "Future Me High Five"
I wrap it up with:
One thing I did last week that future me will thank me for (e.g., finally making that doctor's appointment, not responding to that passive-aggressive email)
One tiny thing I can do this week to earn the next high-five.
It's wholesome. But also motivational. Like if a TED Talk and a dive bar conversation had a baby.
One Thing of Note
Whether you use ChatGPT or Claude to generate a prompt does change the vibe — even the personality.
I ran a little experiment: I asked both ChatGPT and Claude to write a report about me. Then I anonymized the results and compared them. Here’s how they stacked up:
Tone & Style: Claude was more fun and memorable.
Depth of Insight: Claude dug deeper, hit harder.
Actionable Advice: Claude again — concrete and creative.
Narrative Clarity: ChatGPT had the edge here.
Emotional Intelligence: ChatGPT felt gentler and more compassionate.
Brutal Honesty: Claude pulled fewer punches.
Personally? If you want a no-nonsense, funny best friend to coach you through your week, go with Claude.
If you prefer a calm, reflective mirror that helps you see your patterns with warmth and thoughtfulness, ChatGPT’s your pick.
But if you’re chasing growth and can take a little sting? Claude wins. It’s sharper, more emotionally honest, and more likely to push you into action.
Why This Works (Even When I Don't)
It makes me laugh. And if I'm not laughing, I'm probably not doing it.
It gives me insight without the shame spiral.
It turns my chaos into something I can actually work with.
It reminds me that healing doesn't have to be Instagram-worthy. Sometimes, it's messy, awkward, and fueled by snacks from the 7-Eleven next door.
Want to Steal It? Go Ahead.
You don't have to use my exact prompt (though you can if you want)
But here's the real takeaway:
Don't design your accountability system around who you wish you were.
Design it around who you actually are: beautifully inconsistent, sometimes avoidant, occasionally brilliant.
Name it something that makes you smirk. Inject humor like it's the antidote to shame. And give yourself the space to celebrate progress, even when it's microscopic.
TL;DR (because I get it)
How I avoid becoming a complete disaster:
I journal.
I feed the chaos into a sassy AI prompt.
I get insight, baby steps, and micro-rebellions.
I high-five myself.
I do it again next week.
It's not perfect. But it's honest. And honestly? It's saved me from myself more times than I can count.
👏🏼⛽️💪🏼✊🏼💪🏼⛽️👏🏼