As we enter Leo's 4-year-old summer, a few of the little systems we run on stopped fitting him. I'd been tossing some ideas around, but the transition from school schedule to camp schedule felt like a good time to finally get things rolling. No major lifestyle changes or anything - mostly giving him more responsibility (or, maybe more accurately, making his existing responsibilities more concrete) and a bit more independence. During the school year, between activities and everything else, screens were basically just Friday movie nights, and maybe a bit more the dead of winter. He also knows we don't do screens on school nights. His logic is a little too good because as soon as he realized there were months of "no school nights" ahead, he pounced on it and the screentime requests became constant.

I ran a few thoughts through my agents to try to shape my various ideas, and then thru Gemini (the superior image generator, if you're not already using it). Nothing fancy in this one - no background tasks or automations. I used the context I'd already built up in my agents to create a couple basic prompts to bring the system to life.

This edition: Mom Mode: 4 little summer printables.

Three tokens a week

In trying to figure out what to do about the screentime issue, I knew I didn't want him to necessarily "earn" screentime, but I also wanted a way for it to still have some conditions around it and make it something he needs to make conscious decisions about. I landed on a token system - he gets 3 tokens every week, each giving him a screen time session. He can use the tokens any time after lunch, and only if he's completed his "jobs." Fridays are free days because that's always pizza-and-a-movie night. The tokens reset every Saturday.

So where does the AI come into all of this? Besides helping me crystalize how I wanted this system to work, I knew I needed a way for him to visualize his jobs to make them more concrete for him.

Task 1: The tokens.

I 3D printed these. I put a super basic one-line prompt into Gemini, made a token in about 5 mins in Fusion 360, and sent them to the printer.

Task 2: The job chart.

I find the process of writing prompts, especially for image generation, to be really tedious and error-prone. I told Leo's agent what I wanted and she wrote the prompt. 2 minutes later, the sign was waiting for me in the printer (the paper printer this time).

The system comes with the added bonus of working in some basic budgeting concepts. Week one he used up all of his tokens by Monday night. This week he's only used one so far, on Sunday. It's taken so much negotiating about screentime off the table. Highly recommend.

The summer bucket list

A few years in, and I'm still not used to planning ahead. I've always operated on "wherever the weather, flight deals, or whims take me." I'm realizing now that things take a bit more forethought since we're working with a school year schedule AND camp is already locked in. I've always kept a little note with seasonal "bucket lists" of things I thought might be fun to do (apple picking, ice cream places, egg hunts, the stuff you either forget in the day-to-day of life OR need to book ahead), but this time I wanted to get Leo involved. I debated just using a blank piece of paper and drawing on it, but free-drawing isn't QUITE where we are yet, so I made a template for us to fill in:

This is another one where I enlisted Leo's agent to create the prompt and then I stuck it in Gemini. It's our project for this weekend, so we'll see how it goes.

Evolving play = evolving storage

Pretend play is getting more and more elaborate around here. His room is pretty small, so I needed to make it easier for him to not just have space for all of these elaborate stories, but also make it clear how to undo it all when it was time to find his floor (and the living room floor. and the hallway floor.) again. This is one I've been meaning to do since before Christmas, but I kept putting off because I knew a label maker wasn't going to cut it since he can't read yet. I finally decided to make life easier for myself and just print it on paper and use packing tape to "laminate" the labels and put them on his drawers. Agent to the rescue again to create the prompt and stick it in Gemini.

The prompts

They’re all copy-paste ready, just swap in your kid’s details. Gemini or ChatGPT would be your best bets for image generation. If you’ve figured out how to make it work in Claude, please tell me. Claude is truly awful at image generation. They all still tend to over-index on “make a poster” creating an image with a poster on a wall rather than something print-ready, which is why I lean on the agents to make the prompt to begin with. Tell it exactly what you plan to do with the output (print it on X side paper or for a newsletter thumbnail, etc), so you can hopefully at least get close.

In the end, I know he’ll outgrow these systems too, but for now the days are smoother. The screentime negotiations are non-existent (so far) and I'm pretty confident the toys will be a never-ending battle, but at least we have a goal to shoot for, and it’s more context in my agents to build on top of for the next thing.

I'd love to hear what sorts of AI you're using to kid-related things! And if you use any of these, let me know how you improve them!

At least once a week, Leo comes home with something in his backpack that has a date to remember. Then there's the PTO events, the local community events, the random things I find on instagram or facebook. The texts about parties. The email confirmation for the Lowes workshop. I'm a screenshotter, which is great until I forgot I took the screenshot and never load it into my calendar. And if it's not in my calendar, it doesn't exist.

This edition: My screenshot -> calendar skill.

This one is dead simple. Realistically it shouldn't require AI. But again, my goal is to hand off where I'm failing at executive functioning. The number of times a week (or even sometimes a day) that I have to add something to my calendar is approaching comical and I'm coming to accept that anything that can be "set and forget" is a good task for my squad of agents. I made this skill back in March while I was walking to Disneyland from the hotel and it's been paying dividends ever since.

Now, I take an image, send it to my agent, and everything I could possibly need about the event is put in my calendar according to my pre-defined rules. It's one of those little magical things that just makes my life easier.

The secret sauce

Just adding things to a calendar is cool. That, in itself, would save me a bunch of time. But where this skill works best for me is the rules I set around how to handle events so that they hit my eyeballs at the right time. A school event on my calendar at its actual time is useless to me. "Picture day, 10:00AM Tuesday" - Leo is already at school at 10AM. We're well past the point of being able to do anything useful about it.

The school day rule: if it's a school-day event, don't put it on the calendar at the event time. Put a 15-minute heads-up at 7:00AM that morning instead, with a 🎒 on it, and write the prep in the description. Dress him in green and gold for spirit day. Bring a toy that starts with the letter R. Walking field trip - comfy shoes and layers. It comes back to me while I'm in getting-ready-for-school-mode, at the exact moment I can do something about it before it slips my mind.

A couple of other rules I gave it, because the defaults annoyed me:

• One reminder, 60 minutes before. Never the 24-hour-ahead one - that just gets swiped away and forgotten. An hour out is when it’s useful. If there's driving involved, I try to have it account for that in the reminder (this part isn't foolproof yet).

• If I say "remind me to..." it makes a lightweight reminder, not a full event - 9AM on a weekday, 8AM on a weekend. 15 minutes. Reminder 10 mins before.

None of this is fancy. It's modeled after how I function. I trained it once, and I never have to think about it again. Now everything that needs to be put on my calendar goes from a 2 minute task to a 10 second task. It's the difference between a system I actually keep using and one whose novelty wears off by day 5.

How it runs

My version runs through one of my OpenClaw agents - the same always-on setup behind my meal planning. When a flyer shows up, I text the photo to my agent on Telegram, it reads the image, writes the event to my Google Calendar, and texts me back what it did. If the new event lands on top of something I already have, it flags the conflict so I can decide what to do, if anything.

You don't need any of that to get the same result. If you use Claude with skills, this is a skill plus a calendar connection - set it up once and every image you send gets handled this way. It's how I originally set up the skill. If you don't want a skill at all, drop a screenshot into a regular chat with any AI that can see your calendar and tell it to make the event. The agent version just means I never re-explain my rules.

Bootstrap

I put together a starter version if you want it. It's got the skill and the preferences written out so you can see how the rules are structured, then change them to fit your life. Maybe your kid's school does early dismissal and you want a different heads-up time. Maybe you don't have a kid and you just want fun event posts to stop dying in your camera roll.

Download the bootstrap.

Same as last time: I highly recommend letting the AI interview you about your own quirks instead of trying to write them all out cold. You'll forget half of them until it asks.

If you have any small but mighty skills set up, I’d love to hear them!

Everyone is writing about AI. I guess I am too.

I've been looking for a place to document my adventures with AI. I've written this about a dozen different ways. It all felt like noise in a noisy space. But what the hell - maybe it'll be useful to someone.

I got hooked on OpenClaw about a month and a half ago. It wasn't until a few weeks into it that I realized my executive functioning was working so much better than it usually does. Is it because I offloaded a bunch of it to the robots? Maybe. Is it placebo? Probably that, too. Whatever it is, I'll take it. I haven't felt this organized in years. I'm not particularly into all the AI-is-going-to-eat-the-world hype. But if it means I'm not panic-ordering Doordash at 5:30PM every night because I can't figure out what to do about dinner, I'm here for my new robot companions.

This edition: Meal planning.

I can cook. That's never been the problem. I used to excuse not cooking as "why am I going to bother if it's just for me?" But now with Leo, that's not really a valid excuse and I want to be putting more healthful things on the table. I actually love to cook. What I don't love is needing to figure out 4 meals a day, every day, for the next ~18 years. Not to mention I'm a pretty (very...) picky eater and an even pickier cook. I absolutely refuse to touch raw meat, even though I don't mind eating it in some forms. The amount of nuance to my eating preferences is enough to make you give up trying to feed me by the 4th bullet point. I'm trying my hardest not to pass this along to Leo. Every meal is an exercise in threading the needle between what I want him to eat, what I'll actually eat, what I have on-hand, what time I have available, and probably 5 other variable factors depending on the day.

Leo needs a packed school lunch every day. He also needs a packed snack. Tuesdays he has gymnastics and there's no time to cook. Fridays are pizza. A protein I cook on Sunday can't show up again past Tuesday. I don't want either of us eating the same meal twice in a row. I need to eat something other than a handful of pretzels or goldfish for lunch. And then there's the long list of foods that never appear in my kitchen, including meat with bones, mushrooms, and eggplants.

From the dawn of my use of AI, my first ever use case was meal planning. I knew exactly what I wanted a system to do. The rules were clear in my head. But I could never find a system that worked once I got it out of my head. I've tried to set this up before, and it never stuck, because every version still waited on me to start it. And see previous executive functioning comment. I was never going to remember to start it.

Then came Claude's Routines (which is now Scheduled tasks). Then came OpenClaw. Once I had an agent that could run the plan on a schedule, and the week's meals and the grocery list started arriving without me asking, and the whole system finally came together.

The System

My meal planning system is 4 files: Preferences, Inventory, Shopping list, Skill.

Preferences: This is where every constraint lives. They're specific to the point of being almost absurd. A sampling:

• Weeknight dinners cap at thirty minutes of active cooking. Tuesday's dinner is whatever reheats fastest, because there's no time to cook on a gymnastics night.

• Every Monday-through-Thursday dinner has to leave enough leftovers to fill Leo's lunch box the next day.

• A cooked protein has a two-day shelf life. Sunday's chicken can power Monday and Tuesday, then it's finished.

• Leo's lunch needs a protein, a carb, a vegetable, a fat, a crunch, and a fruit. His snack has to survive an unrefrigerated backpack and can't be a single sad item, so two things minimum. Never trail mix, which is basically candy.

• No chicken thighs, for anyone, ever. No raw tomato in my portion, zucchini is fine for Leo but never for me. I will never make 2 different meals at a time.

Pantry: A running snapshot of what's in my kitchen right now, freezer and dry goods included.

Shopping list: Organized by store.

The Skill: the instructions that read the other three, turn them into a week of meals, and spit it out in my preferred format.

5th Bonus File: "keeper" recipes and favorites. Links or other notes for things I make often or I've liked. It pulls from this but isn't limited to this.

Every week, on Sunday, OpenClaw triggers the skill to plan. It reads the rules, checks what I already have, and builds the week around the food in the house. Then it writes the shopping list. It also checks my calendar and the weather to make sure it's proposing seasonal, weather-appropriate options and works around any events. If it's nice out, I'm grilling. If there's an event on a given night around dinner time, it'll flag that and propose something quick or a meal from my freezer stash.

The Output

Every Sunday, without my asking, next week's plan lands as its own dated file and the shopping list updates to match. Each day has Leo's breakfast, his packed lunch built from the night before, a backpack-safe snack, my own ten-minute lunch, and dinner. Friday says pizza for dinner. There's one thing to batch-cook on the weekend, and the groceries are grouped by where I buy them.

And if there's something I likely wouldn't know how to make, a brief recipe is included at the bottom.

How it runs

My version runs through an OpenClaw agent I keep running in the background, so all of this happens on a schedule instead of waiting on me. This could run just as well on a Claude scheduled task. If you're capable of doing it without a scheduler, then you could also just throw the files and prompt into regular Claude Chat, ChatGPT, or Gemini, and get what you need. My OpenClaw agents sends the output to Telegram, where I go back and forth with my agent to refine it.

It's still just a robot

While the skill is pretty good at this point, there's still things it flakes on. And I don't follow it to a tee. But having the baseline there as a backup plan is immensely helpful. I have a second automation that runs every evening that checks in on what we actually ate so it can track the plan vs deviations from the plan as well as keep my inventory up to date. Obviously it's not perfect at maintaining inventory. Be sure to remind it of its rules, and tell it occasionally to review the conversation and write back any rules/preferences/findings that it hadn't previously captured. Make it learn. Make it work for you.

Bootstrap

I completely rely on being "interviewed" by the AI to get the details I might not think of into my skills and files. The bootstrap files will run an interview for you. Or you can just braindump off the top of your head. I highly recommend Wispr Flow (referral link) for the interview process. Speaking freely works much better than writing it all out, IMO. From there, you can decide how it runs: paste them in each week, or, if you use Claude with skills, let a scheduled task do it for you.

I put together a little bootstrap setup if you'd like this for yourself. Have at it, and make it your own. I'd love to hear if you make any cool modifications!

Download the bootstrap.

So I guess this year’s story is just going to start to write itself. Not to bury the lede: I am presently funemployed.

It’s kind of funny, really. I feel like I spent the better part of the start of this year knowing there were layoffs coming at work and trying to make sure people who might not have been through this before in their career were prepared. Always be prepared to be laid off. In my last 1:1s, when RIF-gossip was at its peak, and folks were looking for reassurance that they were doing high-value (read: safe) work, I said over and over again: “Yes, you are. But remember: You could be the objectively best person ever at your job, but if the business decides it wants to move your job somewhere else, then it’s going to do that.”

At least I was taking my own medicine. Two weeks ago, when I was gearing up to lead my team through RIF day, an email appeared in my personal inbox: “Impact to your role.” I was out.

It wasn’t a complete surprise. I’d sensed something was up in the couple of weeks leading up to it. But it was still a surprise in that I’d been told I wasn’t impacted. So it goes. My whole hemming and hawing about what I want to do with this fork in the road? Well - the fork made the decision for me. And if I’m being honest: I’m kind of relieved.

I’m still struggling with the whole identity crisis of it all. My whole career has been tied up in tech. But the idea of going to another tech company just doesn’t feel at all right at the moment. Maybe that’ll change with time, but as it stands, I don’t think it’s my right next move. It feels weird to say out loud the thing that I’ve only said jokingly in the past: I don’t want to do tech as my full-time thing. I want a real, tangible small business. Like an ice cream shop. Or my farm. I want the people I work with, and for, to be real, normal everyday humans. Not to say tech people aren’t normal humans. But we exist in a whole distorted reality.

The plan for now: I’m open to some consulting, advising, teaching, or fractional work. I’ll revisit the whole “Do I go back into tech full time?” conversation toward the end of the summer. For now, while the kid is still little, owning my own schedule is important to me. Without being strapped to my chair in front of Zoom for 8-9 hours a day, I now have the freedom to do the more important things like take him to swim classes or go to the playground after school. I don’t want to trade that right now. I’m going to use my “working” hours to learn things, experiment with things, and, if I’m lucky, find an ice cream shop to buy. I’m also keeping up with A Tiny Workshop and might pick up some vendor events as we enter that season.

As it turns out, Scatterfocus might be the right name for this newsletter after all. Stay tuned for what’s next. Your guess is as good as mine.

Until next time…

Jenn

P.S. I am very open to ideas or suggestions for how to spend the coming months, including (kid-friendly) travel, ideas to explore, books to read, people to talk to, etc. Please send anything my way! 🙏

I’ve gone through many versions of “oh shit. now what?” since the start of November. I researched what it would take to move out of the country. I considered moving back to California. I cried. I lived in a base sense of dread and “why bother” for a while. I considered moving out of the country again. And yesterday—Inauguration Day—instead of vegetating in front of the tv watching terrible Netflix as I’d intended, I hid myself in my unheated basement, anxiety-ordered a new 3D printer, and got to work on revitalizing my Etsy shop. It wasn’t until a few hours later when I was scrolling through my shop’s instagram feed (specifically avoiding my personal account and any other news headlines or social media) that I saw a couple reels on the theme of “accidentally protecting your peace too much” that put it all in perspective for me. I was, subconsciously, protecting my peace.

A few weeks back, I wrote what I thought was going to be my next newsletter entry about reflections on 2024 and goals for 2025, but something wasn’t clicking. I was trying to force myself into some theme for the year or to find some thread to run through it all about financial independence or figuring out the next few years of my life. But, again, nothing was sticking. There’s too much ambiguity in life at the moment, in both the macro and micro senses. Setting some sort of decision-making framework around what gets me to the point of semi-retirement or a career break or whatever it is the future holds, feels too big a decision with too many existential unknowns at the moment. I couldn’t get it to come together. Until those reels.

My operating model for this year is going to be “How can I best protect my peace?” It’s probably going to include staying off of the news, which I did for a while, but the glimmer of hope between July and November sucked me back in. It’s probably going to include ruthlessly curating my social media feeds (I closed my Threads account and moved to Bluesky if you’re there). And, for now, it involved me muting a group chat with two of my oldest friends because they are ok playing in the snark zone of dread and I’m just not there.

I know I can’t hide my head in the sand from what’s going on in the country/world, but I do need to be OK with blocking some of it out and leaving room in my goals and plans for existential uncertainty. I still have a goal of figuring out a plan for semi-retirement or whatever’s next. Making those sorts of decisions right now, when it’s unclear what the economy is going to do, feels too risky for my appetite. The side projects that give me a creative outlet and end-to-end control have always been a grounding element for me in uncertain times. 3D printing and hanging out on maker/small creative business Instagram is such a wholesome place compared to the real world. It feels like what I need right now.

All that is to say: I’d dusted off this newsletter to explore what it might mean for me to “retire early,” but I originally started it as a documentation of whatever hobby or topic I was pursuing at the moment. Looks like the pendulum, for the moment, has swung back to where it started. So for now, I guess it’s kind of a blog of miscellany? I hope to provide some more “valuable” content in the future.

As an offering, here are some of the ways I’ve kept sane in the last few weeks of seeming-impending-doom mixed with a two week bout of daycare plague:

• 📺️ This series “Inside Everest” from Business Insider. It’s about a year old, but super interesting.

• 📺 I found this guy Liam Thompson’s channel, and I found his ability to make somewhat mundane topics into whole, long-ish form, engaging videos really interesting . Especially the Shakespeare one which is literally a video about him reading a book.

• 📺 Speaking of mundane, this one about “fire before electricity” was kind of fascinating. Thinking about candles as a technology is interesting and I was shocked by the part about innovation on wick braids.

• 📺 This video on burnout hit more than most articles I’ve read over the years. It was before my whole “maybe I should just hide in the basement and 3D print for a little while” revelation, or maybe it played a part in it.

• 🏗️ Magnatiles. Leo is finally into building and can easily lose an hour or three to these. I personally find them infuriating because they don’t physically connect and fall down super easily, but they do have a very soothing clack sound.

Until next time…

Jenn

Oh, hello there. You might not remember, but sometime since mid-2022, you signed up for my newsletter. It’s been collecting dust in a corner of Substack, but I set myself a “resolution” to try to find the older, smaller internet again. So here I am, and on beehiiv this time - Substack seems a bit controversial these days. If you don’t remember, or don’t care, you can unsubscribe. I won’t be offended.

I’ve been stumped at what to write for months now. According to my ChatGPT, I started putting this newsletter together in October. It started with me downloading the archive of my blogs from the dawn of my blogging days. I uploaded it to ChatGPT and asked it what I used to write about. As a surprise to no one who knows me, the theme was just “whatever I was focused on at the time” - a mix of travel, tech, work, life, projects. Pretty much the same as I am now, just with more time to write about it.

Like many of you, ChatGPT has become an outsourced brain and processor for me. It’s come to know a fair bit about my various interests. So, if I’m being honest, when it came to finally sitting down and writing this thing, I asked ChatGPT “based on what you know about me, what could I write a newsletter or blog about?” Turns out, it’s all of the above. With some financial independence, parenthood, and product management mixed in. When I asked it the common thread running through my interests it spit out: "Designing a Life of Intentionality: Balancing Leadership, Creativity, and Parenthood While Building Toward Freedom." I think - or at least I like to think - it’s pretty spot on. Yes, my interests are kind of all over the place, but they do all point in one direction. And that direction is one that I don’t think I’ve ever quite shared publicly: My driving goal in life, since my teens, has been to retire by 40. 

If you haven’t already, prompt ChatGPT with “Based on what you know about me, draw a picture of my life.”

This year is one of those fork-in-the-road years. I have some (though vanishingly short) time before 40, and technically I have reached the financial number at which I could retire. But I don’t feel mentally, or logistically, ready yet. So my plan is to use this space to explore that to do list. And maybe do some unpacking of why I feel so scared to actually do it (spoiler: if I had to guess, it’s society, norms, pressure, expectations, and all that crap).

Here’s the list of things I’m noodling on as it stands now, in no particular order:

• What would soft retirement look like? I don’t want to be done working forever. I enjoy having some sort of “useful output.” Retirement looks somewhere between “I don’t have to continue to work a high-energy tech job anymore” and “I can open up that pop-up ice cream shop / bakery / co-working space / alpaca farm I’ve always wanted.” Time freedom is what’s important to me. Time to explore my interests, do fun things with my kid, travel, and not be attached to Zoom meetings 9 hours a day.

• Do I want to continue to live in this house? There are pros and cons. Considering school district, space, income potential, etc. it might make sense to move somewhere else and use this as an income property/summer house.

• Should I pick up some consulting/fractional work for a bit to ease out of working full-time in tech? I really enjoy what I do, and I think part of what’s keeping me from stopping is I don’t want to give it up entirely.

• Should I find a property to buy, renovate, and use as an income property (or flip?). I love renovating. That would be a fun combination of hobby-and-investment.

• Should I double down on my Etsy shop? It’s not nearly as successful as it was during that flash-in-the-pan spike of Holiday 2022, but I do still get orders with zero marketing effort. And again, I enjoy it.

• I’ve always loved to teach. I currently teach Product Management through Kellogg’s Executive Education program. Though, given the state of the market, It hasn’t been consistent as it used to be. What other opportunities are there

How can I give myself some structure to be able to do these things or test the waters before taking the plunge? This used to be a lot easier when I had nothing but time on my hands. But with a 2.5 year old, free time is almost a joke 😆. Though, I am excited about the idea of getting him involved in some of these projects.

I’m hoping that by putting these thoughts out there, somewhere, I’ll be able to find that small internet again. That old internet where multiple friendships and jobs came from comments on my blogs. Where I would learn about whole new things from trawling Twitter threads or scrolling my RSS feed. Where ideas were shared and feedback was given in a way that didn’t involve a war with comment trolls. So please, if this is of any interest to you, drop a note back! Or share what’s worked (or not worked) for you. As much as I loved yelling into the void on social media, I’m hoping that someone will start to yell back again and we can all learn from each other.

Until next time,

Jenn