I have a hard time starting personal projects. A lot of people I know struggle here and based on some cursory research it’s a pretty common struggle.

This month I decided to start painting again. It’s a hobby I love but I’ve managed to put it off for a few years. Even after committing to it, I debated backing out before my first painting. I had a lot of anxiety about how rusty I’ve become and whether or not I really have time for it.

I also decided to start on some landscaping projects that I’ve been putting off for about two years. I felt a little more sure about this one but I’ve put it off for so long due to lack of planning and time concerns.

These both represent common shapes of the problem for me.

Why is it hard?

Starting requires a change in state.

You need to go from not doing to doing. Seems obvious but there’s a lot buried in there.

It’s easy to keep things as they are and people tend to be resistant to change, especially large change. Starting is change. You have to context switch and build momentum from nothing.

I was resisting getting started with painting this month. It was mostly about the cost of getting started. I had to dig through some boxes and find my paints and brushes. I had to clear off a desk to use it for painting. I had to set up some lighting so I could see what I was doing. I had to actually set aside time to do it.

But then, once you get past the initial speed bump it’s natural to keep going because, again, we are resistant to change. Stopping also has a cost.

You must cut through ambiguity.

Before you start, it feels like a ton of work because you don’t really see the specifics.

We often think in terms of broad, non-specific activities. I wanted to start “painting” again. I wanted to “do my landscaping projects.” You see some ambiguous idea of an action and that can be anything.

Getting started means getting specific. And that is work. You have to decide exactly what you’re doing and not doing. You have to figure out how you’re going to do it.

Once you actually get specific you can bread down the task into simple steps. Take some shovels of dirt and move them over there. Take some paint and put it on that paper.

Ideas can feel perfect. Reality can not.

Behind the other two is the painful reality of, well, reality.

In our minds we can dream up whatever we want. My landscaping projects can be like the Hanging Gardens of Babylon. My paintings can be like the melding of Andrew Wyeth and J. M. W. Turner. Yea, reality isn’t like that.

This kind of compounds with the other two. As we dream of all the ways our work can be perfect, we make starting even harder. The idea becomes larger and more ambiguous. The cost of starting becomes harder to pay.

So, it’s easy to fall into a state of dreaming about something you’ll do and then never actually doing anything.

What can we do about it?

Realize the problem is a feeling and not a reality.

The biggest hurdle to getting started is the feeling that getting started is hard. That’s what we’re trying to address. That feeling can be totally overwhelming in some cases.

But, with this in mind and a few tricks you can make almost anything easier to begin.

Make the first step stupidly small.

There’s a pretty common trick called the “5-minute rule.” Not sure where this comes from but the idea is if you are having trouble starting something, just commit to five minutes of doing it.

This is kind of a hack for picking something small enough that you can do it in five minutes. It might work for you, but I’ve personally never found this time approach that useful. Instead I prefer to directly choose something specific and tiny.

Part of my landscaping project involves moving about 37 yards of dirt from this huge raised bed to some other places in my yard where concrete was removed. I put this off for whole years now. To start I simplified this to just go take a single shovel of dirt and move it where it needs to go about 10 feet away. Now I’m slowly making progress on the project (though I did end up getting an excavator).

Lower the stakes.

Since we like to dream big, we tend to raise the stakes of the outcome. This creates a lot of psychological baggage that keeps us from getting going.

Instead of focusing on some perfect finished product, you can focus on some simple low-stakes intermediate product. Or, just lower the target and plan to improve over time.

For my watercolor paintings I decided to do them in a sketchbook for a specific reason. If I make them individual pieces on individual pieces of paper then I feel like each one could be “a work of art.” But, I’m not trying to make works of art. I’m trying to practice painting and relax.

Putting them in a sketchbook lets me focus on filling the sketchbook instead of creating individual meaningful pieces. Now, I can hear some voice saying “but what if you mess up a page?” I can see how that would bother many people. It might even bother me, but my goal is to fill it and a messed up page gets me there. You have to find ways lower the stakes that work for you and your own peculiarities.

Make a commitment.

You need to actually commit to the thing you want to start. This may sound like it could make things harder. And yet, I’ve found this to clear one of my biggest hang ups.

I have all these things I want to do. But, when they happen can be just as ambiguous as what is going to happen.

For example, I put off these landscaping projects for years. This wasn’t because they are particularly insurmountable but because I never committed the time to do them. I could move some stones around one weekend. Do a little design a couple weeks after that. Then, suddenly it’s a few months later, and I’m doing some research and planning.

No, that’s not working.

Right now I’m actively choosing monthly things to focus on. This month, landscape painting and landscaping. Next month, we’ll see (but probably going to involve more landscaping).

It requires deciding both what will be the focus for now, but also what will be pushed off for later.

Give yourself an exit.

Again, part of the ambiguity is around time. Is this thing going to go on forever? How long will it take?

We address this by making the steps smaller, getting specific, and lowering the stakes. But then there’s still this still this fear of commitment to something new.

It’s always good to have an exit plan. It’s the other side of bounding goals.

I’m committing to painting 30 paintings this month. That won’t even fill my sketchbook. But, if I feel like it I’ll keep going next month or in the future. It doesn’t have to happen now. For my landscaping projects I have a couple milestones that I’m trying to hit this month but if it’s not going well I’m willing to eject and hire people.

We need to design around our emotions.

It’s tempting to think that getting past this problem is about willpower. Suck it up. Just make yourself do it. But, I don’t think that’s the real solution. I don’t even think that’s very productive.

Even if you pull off some force of will, you might end up making yourself hate something you could otherwise love.

The real solution feels more like a design problem. We need to focus on the right problem, the feeling, and then address it thoughtfully. We have to create rules and context for ourselves that’ll make the feeling of starting more positive and less negative.

It’s less about force and more about care.

For the past month, I’ve been publishing articles every day.

I wasn’t really aiming to grow my Substack. Even after figuring out how Substack works I decided to continue publishing these into the void.

The goal

My real goal was to get a feel for writing regularly. Not just writing, though. I wanted to write things that are meant to be read. I wanted to figure out where I needed to improve and what I enjoy writing about. Last month I started doing daily “morning" pages” which have a very different purpose.

The problem

For a while, I’ve wanted to start building some kind of online presence. I feel a need to do this for a variety of reasons. A little bit safety for my career and a little bit opening other doors. Perhaps mostly, I have a lot of thoughts and perspective that I hope could be useful to others.

But, I kept stalling out on this, trying to answer all the questions before I started. This is a very bad idea and has a tendency to lead nowhere.

Lowering the stakes

While writing morning pages one day, it occurred to me that I should just crap out a bunch of articles for a month. I was already writing garbage for no one. Wasn’t much different to write for someone even if no one reads it.

And, it’s better to write trash than to write nothing.

Reframing the goal so that you lower the stakes is powerful. It helps you convert overthinking into action. The barrier to entry drops. Action becomes the success state instead of perfection. Inaction becomes failure. It’s hard to improve when there’s nothing to improve but we can always improve what we are already doing.

What did I get out of this?

Not an audience. That’s for sure.

Committed focus is a strategy for repairing shattered focus.

This idea that I can take a month and focus it towards some goal was the biggest takeaway for me. I’m one of those people with a hobby graveyard and too many things to do. Not that I have to do, that I want to do.

I’ve tried so many strategies for dealing with this. Time blocking calendars fills my calendar with things I’ll never do because the stack collapses. Just going with the flow goes to me doing random shit and mostly nothing.

Committing to extended periods of focus is where it’s at for me. I’ve done it before as proof that I could learn to paint. This one started in a similar way.

The idea relates a lot to hobby seasons. As I went through the month I started to see all the benefits that can come out of it. I feel more control even if I really have no specific plan. Reframing the goal from quality to quantity makes the goal feel achievable.

I plan to keep experimenting here.

I need to protect my time.

All that being said, I can’t keep adding daily practices every time I do one of these. That would be unsustainable. I feel like I want to keep some daily writing practices but I also want to try other things like getting back into painting.

There’s all this stuff I like doing and I have very little time to spread between them. That’s what lead to this structured focus in the first place. Having too many daily practices cuts time for each dramatically.

I’m not sure what the ceiling is though and I’m going to feel it out.

It reduced my social media consumption.

Focusing on creating reduced my YouTube habit. This also happened when I did a similar challenge years ago. You don’t have the time to consume because you’re working on your own stuff. The hidden benefit behind that is I don’t even miss it.

Ideally I could find a way to pull historical numbers on this but the YouTube app doesn’t appear to go back further than a week. Most days for the past week are 0 with just one being over an hour. I was listening to lofi while doing something mindless at work.

My average in the past has been well over an hour per day. Last week it was 25 minutes.

I have a clearer sense of what I’m doing here.

I found that I think a lot about taste, identity, AI, attention, product and software development. But that’s a little all over the place.

Going forward, I’m going to focus on exploring these monthly hobby seasons and goals. Along with that I’ll focus on topics like attention, identity, and taking action. I’m going to stay away from topics like AI, product, and software development. Taste falls somewhere in the middle. It’s an aspect of your attention and identity that affects what you make and curate.

I do have a lot of thoughts on digital product development but I’ve never released a successful product. That’s a cold reality that I struggle with. So unless I’m positioning my thinking around what not to do (which I know a lot about) then I don’t have a lot of advice there. Certainly not advice that you couldn’t get from someone else.

I’m also hesitant to give real world work examples because, well, many of them would be considered failures. It would feel like blasting failed projects I’ve worked on with good people. I’d rather blast myself, if anyone.

Daily publishing is too much for me.

For the first week or two my articles were mostly just brain dumps of things that were top-of-mind. Those were pretty easy to write. After that I hit a slump where I felt very much out of things to write about.

Some advice says you need to keep publishing daily to grow but I don’t buy it. There are plenty of people who succeed at writing with much less frequent publishing schedules. Plus, I have a full time job, a family, and other things to do. That’s just not realistic for me.

Writing daily is great.

On the other hand, writing daily gives me a lot of value in the form of just getting things out of my head. I had already learned this from the morning pages. But morning pages are different than writing about something.

Morning pages are pure stream of consciousness that sometimes turns thought provoking. It has this beautiful emergent and clarifying quality.

Daily essays require me to try to pick a point and make it. This is more painful as I sometimes feel stumped. But I’ve gotten a lot of value by trying to routinely solidify my thoughts.

This painful point where I ran out of pre-built steam is where I started to find ways to come up with things to write about. I dredged old chats with friends and AI. I took two ideas that I felt were related but couldn’t quite see how and tried to connect them.

I’ve long been a fan of the idea of writing to learn. It’s this point where you’re following threads and connecting things for yourself that it actually happens.

Substack takes work beyond writing.

I was totally new to Substack when this started. I’d made an account a while ago but did nothing with it. I noticed as I was going that no one except maybe the couple friends I have subscribed were reading my posts.

So, I did some research and learned that I need to do more work to promote my writing. I assumed somehow an algorithm would take care of it, like many other social media platforms. I can’t say I’m particularly thrilled about this but I’m going to give it a shot and see what happens.

What’s next?

I’m going to continue experimenting with these monthly seasons.

I have many ideas for them already. Honestly, I could probably list 50 of them off the top of my head. That’s another kind of proof that they work.

All these ideas and goals realistically need large chunks of time. Focused, dedicated time used to move them anywhere. Giving large hobbies and goals little spurts of time here and there goes nowhere.

Committing and limiting other commitments around them gives them the space that they need. It gives a sense of control and a way to make trade-offs. It helps me think about what is really possible in a year.

With that, the hobby graveyard becomes a hobby garden.

I’m not really a fan of social media. I’ve been cutting it out of my life for years now. Facebook, then Reddit, slowly YouTube. I’m tired of it sucking up all my time. Though, I actually work at a social media startup but that’s a separate bit of irony.

Why am I here and what’s the problem?

I feel the need to have some kind of online presence and I like writing more than editing videos. Looks like I’m not alone in this hate-need relationship.

To get started I made myself write on here every day for a month. I’m nearing the end of that but noticed that no one reads my Posts. All things considered, I don’t find that particularly shocking. So with that point of view, I brushed it off for a while.

Then, out of curiosity, I started looking at Substack a little more closely. At first I was just treating it like a blogging platform. I came into it knowing it’s for newsletters but I saw there are other social features and basically treated it like Medium expecting people would be able to stumble upon my posts without me doing anything. Turns out it doesn’t quite work like that.

At some point Substack seemed to notice that I was confused. I never asked but The Algorithm started sprinkling my feed with “how to use Substack” type posts alongside wonderful pictures of trees and art. So, I bookmarked a bunch of them and started reading.

I’ll leave a list of posts at the bottom.

Without further ado, here’s what I learned about growing on Substack.

Do you want to grow? Why?

The thing you need to decide up front is what you’re trying to accomplish here. I came into it unsure and just started writing. That’s probably fine too.

But, depending on your goal you may need to approach Substack in different ways. Some people are trying to grow free subs and some people want paid subs. Some people are trying to build a funnel and some a community. There are many possible goals. All of them require different strategies because there are different audiences.

Know who you’re writing for and what you’re writing about.

There are a lot of different types of audiences you could be trying to reach. You need to know who those people are. Perhaps more importantly, you need to know how you’re trying to help them.

This interacts with the types of growth you’re chasing. Are you trying to give them strategies that make them more money? You should probably get paid subscriptions. Are you trying to offer something off Substack? You might want to use free subscriptions for that. Are you just wanting to write on an interesting topic? That could go either way. Maybe you should fall in the middle and add a “Buy me a Coffee” button.

You should have a niche but not one that’s overly narrow. You need to have enough people that you can find them. The niche should probably be around the problem you solve or how you help people. But, that’s for you to figure out. No one can tell you what this should be.

The most important thing is that you figure out who you’re writing for - because you need to know that to give them value and correctly position yourself.

Make your position clear.

At some point you’ll know who you’re writing for, why, and what you’re writing about. With that, you need to make this clear and easy to navigate. Substack has a lot of possibilities for customization. You should use these to clarify your brand and make your content easier to navigate.

Your position should be clear in your bio, footers, Notes, or wherever people try to figure out who you are.

For example, after you get going you could write Posts specifically used to orient people to your writing and its value. These can be placed on your home page or included in your footers.

Find a cadence.

As you start writing on Substack, you’ll need a cadence. Some people recommend daily, some a few times a week, others once a week. This is really a personal decision but you should pick something that won’t burn you out because consistency is critical.

If people subscribe to you, especially if they pay, they expect regular publications. And the cadence of these publications should be part of what you’re making clear to them.

You can pre-write and schedule Posts, if needed. This can help you work and bursts but still maintain a steady cadence.

Posts don’t share themselves.

This is the thing that I didn’t know about. When I started my Substack I assumed that Posts are automatically shared somehow and could be discovered. This isn’t strictly untrue but skips the primary method of discovery.

If you want people on Substack to find your Posts, you should be promoting them through Notes. This won’t guarantee anything though, it just makes it possible.

On top of that, you can share links to your Posts on other platforms like LinkedIn and Instagram to share with audiences you might already have. You can also set up your Substack so that Google can find it.

Notes also are a general path to discover you which will lead people to your Posts indirectly.

There’s also a lot of discussion around Note recipes and patterns. Some may be more engaging than others. I personally don’t love these kinds of tactics but you may find the information valuable. Generally though, the more engaging, valuable, and sharable your notes are the more likely they’ll help you grow.

You need to actively engage with the platform.

Posting and creating Notes isn’t enough. You should spend some time engaging with other writers, liking and leaving thoughtful comments, Restacking things that you find valuable (or your audience might).

Basically, you should be acting as a member of your community. You don’t need to spend too much time on this but a little time throughout the week can help you in several ways. The Algorithm may favor you a little more. People may reciprocate and share your content or recommend you.

Just like Note patterns, it’s important to do this genuinely and not solely as a growth hack. People can tell.

Don’t worry too much.

It’s tempting to overthink all of this. You could spend ages coming up with the perfect positioning and planning out content. Forget all that and just start.

I decided to just write for a month on here so I could get a feel for it. I haven’t done most of this yet. I’m using these Posts to help me figure out exactly what I want to write about and for who. I needed to feel it so I could figure out what I enjoy writing about and what value I can provide.

You can do this too but you don’t have to publish. You could keep them private and bank the ones that fit your audience and niche for later.

You also don’t need to figure out monetization too early. Figuring out what value you provide is more important. With that, you can start thinking about what monetization looks like. If anything.

Don’t get sucked into all the growth hacking. Be careful trusting anyone that tells you exactly step-by-step how to grow. That could work in some situations and not others. Paid growth is a lot different than free growth, for example. Plus, that kind of optimization is for later, when you have something to optimize.

Just write what you think is good and share it.

Sources

Average Joe’s Guide to Substack

The Not-So-Secret Formula for Substack Notes

Hey Guys…

Read more

a year ago · 863 likes · 161 comments · Frey

Average Joe’s Guide to Substack

4 Things to Stop Being Invisible on Substack

If you're someone who started with zero online presence…

Read more

a year ago · 285 likes · 50 comments · Frey

Average Joe’s Guide to Substack

A step-by-step guide to polishing your Substack (Beginner Edition) Slightly warmer

Quick reminder: My friend Sinem Günel is hosting a FREE Masterclass all about how to grow on Substack A DAY FROM NOW! (January 15th, Thursday…

Read more

5 months ago · 1029 likes · 88 comments · Frey

Write • Build • Scale

13 Ways to Get More Subscribers Using Substack Notes

Read more

8 months ago · 856 likes · 169 comments · Sinem Günel

Ditch The Templates

I Studied 100 Substacks: Here’s How the Top Writers Actually Make Money

Sixteen months ago, I hit publish on my first Substack post…

Read more

10 months ago · 1657 likes · 34 comments · Andi Bitay

Write • Build • Scale

What I'd Do If I Had to Start on Substack From Scratch

Read more

a year ago · 5927 likes · 832 comments · Sinem Günel

Content Orchestration

Do Not Monetize Substack With Subscriptions – Do This Instead

Subscriptions are a fantasy dressed up as a strategy…

Read more

a year ago · 3836 likes · 858 comments · Matt Giaro

Writing Club

The Bare Minimum Guide to Starting Your Substack Newsletter

I see you there, lurking in the Notes and comments sections…

Read more

a year ago · 1201 likes · Writing Club and Anna Seirian

The Second Act Strategist

The Metric Nobody's Really Talking About on Substack Notes

By now you’ve heard the news… is it even a newsflash anymore if you see it everywhere?? 😂…

Read more

2 months ago · 143 likes · 31 comments · Tracy Friedlander

The Second Act Strategist

The Metric Nobody's Really Talking About on Substack Notes

By now you’ve heard the news… is it even a newsflash anymore if you see it everywhere?? 😂…

Read more

2 months ago · 143 likes · 31 comments · Tracy Friedlander

Landon’s Letters

Why Most Substack Writers Are Invisible (And How to Fix It)

Your Writing Deserves More Readers…

Read more

a year ago · 617 likes · 138 comments · Landon Poburan

The Human Project

Your Newsletter Can Make $8000 Monthly If You Stop Following Popular Advice

Most people start newsletters with big dreams, but quit when subscribers don’t show up…

Read more

10 months ago · 559 likes · 45 comments · The Human Project

Sunday Ripley

I'm adding a "Buy Me a Coffee" button and why you should too

I love sharing my writing with you, and I don’t want a paywall standing between us. Stories and posts should be read, not locked away…

Read more

a year ago · 3406 likes · 448 comments · Sunday Ripley

Write • Build • Scale

I Analyzed 9,641 Substack Notes. Here's What 99% of Writers Are Doing Wrong.

You wrote a Substack Note last week. It got 80 likes, a few restacks, maybe a couple clicks to your publication, and zero new subscribers…

Read more

a month ago · 171 likes · 25 comments · Sinem Günel and Orel

Unplugged by Yana G.Y.

I Analyzed my 1611 Substack Notes with ChatGPT — Here’s What I Found About Virality and High Engagement

Many writers complain of slow growth right now on Substack. I did see some slowdown in the last few days as well…

Read more

a year ago · 738 likes · 171 comments · Yana G.Y.

Creator's Playbook

How I Gain 10–30 New Subscribers Every Day Using Substack Notes

I started writing on Substack in July 2024…

Read more

8 months ago · 687 likes · 149 comments · Mark Wils

Escape the Cubicle

I Posted on Substack Notes 365 Days Straight. Here are the only 3 types that brought subscribers.

I’ve posted on Substack Notes every single day for the past year…

Read more

7 months ago · 3215 likes · 411 comments · Wes Pearce

PubStack Success

The Notes algorithm explained (by its actual creator)

This post was brought to you by the Substack Starter Kit, your all-in-one toolkit to launch and grow a successful newsletter with confidence and speed. Everything you need to start, plan, and grow a Substack publication, ready to use and easy to follow…

Read more

7 months ago · 2084 likes · 225 comments · Karen Cherry

The Second Act Strategist

Stop Writing Notes. Start Writing Positioning Statements.

I’ve been serious about Substack for about 6.5 months, and during this time I’ve been actively studying notes that get serious traction — not from the biggest accounts, but from writers who consistently get engagement from the right people. The goal is to learn what works on Substack, because every platform has different things that make it tick…

Read more

6 months ago · 628 likes · 112 comments · Tracy Friedlander

The Second Act Strategist

Stop Posting Notes. Start Posting "Worldview Manifestos"

The more time I’m on Substack, the more I am loving digging into writers with the most successful posts and notes out there. I’m constantly reading, analyzing and taking notes… on Notes…

Read more

5 months ago · 389 likes · 88 comments · Tracy Friedlander

Slay with finances

How I Gained 0-20+ New Subscribers Daily Using Substack Notes (My Proven Strategy)

Most writers spend months perfecting a newsletter nobody reads…

Read more

2 months ago · 170 likes · 46 comments · Jessica Onne

Landon’s Letters

If Notes Disappeared Tomorrow, Here’s How I’d Grow on Substack

It was October 27, 2025, when everything changed…

Read more

4 months ago · 140 likes · 46 comments · Landon Poburan

The Irresistible Writer By Derek Hughes

Substack Isn’t Random. These 5 Patterns Took Me To 7,000+ Subscribers

Read more

5 months ago · 262 likes · 19 comments · Derek Hughes

Really Good Business Ideas

23 Substack Bestsellers Share Their #1 Tip for New Publications

Read more

a month ago · 796 likes · 125 comments · Casandra Campbell, Yana G.Y., CJ Gustafson, Fleur Hull, Ana Calin, Derek Hughes, Adam Ming, Kevin Indig, Tim Stoddart, Xinran Ma, Russell Nohelty, Rich Miller, Jari Roomer, Claudia Faith, David McIlroy, Tom Orbach, Claire Venus ✨, Sinem Günel, Veronica Llorca-Smith, Kristina God, MBA, Ruben Dominguez, Philip Hofmacher, Daria Cupareanu, and Jeff Morhous

Really Rich

How to start on Substack and get your first 1,000 free subscribers

Six months ago, I made the move to start writing on Substack and I’ve made no secret of the fact that I LOVE it here…

Read more

5 months ago · 3346 likes · 555 comments · Helena Di Biase

Move The Needle

How I Went From 0 to 130 Subscribers in 14 Days (And Why I'm Changing Strategies)

I used to put hours into writing, only to get 0 views…

Read more

6 months ago · 914 likes · 263 comments · Charles Daymond

Slay with finances

I Analyzed 100 Substack Writers Making $10K+/Month. 89 Broke This 'Rule'

So I analyzed 100 Substack writers making over $10,000 per month. 89 of them barely engaged compared to what I was doing…

Read more

4 months ago · 1677 likes · 248 comments · Jessica Onne

The Writing Long Game

Have Less Than 100 Subscribers? Here's The 4 Best Ways to Grow

Imagine this…

Read more

10 months ago · 162 likes · 55 comments · Tom Kuegler

Petal + Hearth

The One Habit That Took My Substack from 0 to 21K+ Subscribers

Olivia here! I’m a bestselling Substack coach living in the south of France, here to help you romanticize the everyday and step into your 💫 main character energy 💫 At Petal + Hearth, I share stories and strategies to help you design a life you love, and a creative business that supports it…

Read more

a month ago · 711 likes · 163 comments · Olivia Wickstrom

The Work That Holds

The Golden era of Substack is over. Now what?

The golden era of Substack is over…

Read more

a year ago · 763 likes · 231 comments · Benjamin Antoine

Ambitious Softie

10 things I'd tell you about Substack if I wasn't afraid to hurt your feelings

Two years ago, I was speaking with a literary agent about writng a book proposal. They shared that building a significant email list would dramatically increase my chances of publication…

Read more

10 months ago · 2755 likes · 362 comments · Anna Seirian

How To: Thought Leadership

Problem: Your Substack Is Invisible To Google. Let's Fix It In 10 Mins.

(Even if you don’t have a Substack, read this! You need a framework for what your writing should do for you…

Read more

2 months ago · 1235 likes · 203 comments · Julie Michelle Morris

Personality tests are fun

I’ve taken a lot of personality tests over the years. Occasionally I’ve had to do them for an employer. Sometimes I’ve done it for research. Quite often I do it for fun.

If you’ve never taken one of these, they can be thought provoking. Here are a few I like:

• CliftonStrengths - first took it in college, took it again when I became suddenly unemployed to help rewrite my resume - high quality but isn’t free

• Cake - discovered while researching for a career-building application that I currently work on - free

• Vizzy - discovered during research and not a test in itself but has an interesting one built in built by Clarity4D - free

  Thanks for reading! Subscribe for free to receive new posts and support my work.

I’ve taken (and built) others. Taking one is kind of fun but taking several and seeing how they juxtapose is fascinating.

These tests aren’t trying to find your skills. They won’t tell you that you’re good at computer programming or road construction. Instead, they’re trying to identify qualities that make you you and differentiate you from other people.

This idea is really compelling to me because I think the most interesting aspect of people is their mindset and worldview. Less so what they do with it.

But, our personalities don’t work like this

There is an assumption built into these tests that they can capture a truth about you that is consistent and fundamental. Yet, take these same tests a few times and you’ll probably get different results.

I’ve taken the Cake test four times. I’ve received Tiramisu and Donut each twice. There is some consistency there, of course. Yet it suggests that I am focused sometimes and exploratory at other times. Now, that may be true. In fact, I think it is very true. But how and when does this shift? Can I control it?

In the fall of 2024, after being laid off, I took the CliftonStrengths report to help me rewrite my resume. My results suggested that I tend to be stronger in ideation, adaptability and fixing things. I’m a conceptual problem solver who is very empathetic. I agreed with that and I still do.

It’s kind of in line with Donut.

But then, before writing this I decided to drop the sixty bucks and take it again to see if it would come up different. This result is different in some critical ways.

It now says I’m much more about doing things and individual contribution. I suspect there have been some actual shifts in that direction. But, can I be sure? Or do I swing back and forth between conceptual exploration and acting on that exploration? I think that’s closer to the truth. And that aligns with Cake’s swing.

This result is much less about being empathetic, which is an aspect of my personality that I don’t think has shifted over this time.

Personalities are fluid

These tests rest on an the idea that personalities are a concrete aspect of a person’s identity. This is an idea in pop psychology too. You can change aspects of yourself but your personality is immobile.

There may be some truth in that but personalities are also very fluid. I know this is true for myself. The things I care about and the way I approach the world ebbs and flows over time. My philosophies and values change. There are probably common pillars but a lot flows around them.

A more honest version of these would track personalities over time. By doing that you’d likely be able to identify the common pillars and the ways your personality shifts. Without capturing this element, personality tests plant a stake in the ground that may not be there.

You may be saying “who cares?”

Vizzy uses a personality test as part of a professional profile to hire people. I have taken personality tests as part of an application process. Once I “failed” but was hired anyway because I knew the hiring team. They can have very real consequences.

So, we could make them more honest by tracking personality over time. Then, maybe, they would be useful in hiring. But who wants to take one of these weekly or monthly? Probably nobody. The best ones are long and they aren’t that fun.

I’ve been reading the Daevabad trilogy. Like a lot of contemporary fiction it loves a cliffhanger chapter break. A reveal, a twist, some new insight right before the page turns.

We structure our own lives in chapters too. But we break ours in the wrong places. Resumes, how we introduce ourselves at parties, how we narrate our own histories to ourselves. We love to mark the chapters at logistics. Moved cities. Changed jobs. Had a kid. Fiction knows better. Fiction breaks at the revelation, not the relocation.

My own story could be told like this:

• Did QA automation for an online chip marketing company.

• Took a year off to stay home with my son.

• Did a variety of contract programming work.

• Worked as a software engineer for a car company.

• Worked as a software engineer for an AI company.

• Worked as a software engineer for a career company.

This is how my resume looks if I boil it down.

But several critical things happened alongside these which are far more important to my identity.

I took up painting through a daily practice challenge while I stayed home with my son. This rekindled my love of art and pushed me more into design. I ended up involved with an art show in Canada and started joking that I am an international artist. All of this reframed how I see what I do and helped me connect engineering to art.

I haven’t seen myself as an engineer for years. The exposition was a daily challenge I took up just to prove I could. It taught me something about myself. These events are on my resume but the important revelations are not part of a list of skills and responsibilities.

There’s a subtle reality hidden in here too. This transition of mindset has no clear event tied to it. I don’t know when I seriously started seeing myself as an artist. But I know this helped build that shift.

We need time and reflection to see shifts like this in our lives. We can’t always see our best chapters as they are being written. We can only look back and find them later.

Just because the logistics of our lives are clear doesn’t mean they define us. We get to define our own identities with our own revelations.

One of the many amazing things about LLMs is their capacity for reasoning. With this, we can automate tasks that were practically impossible several years ago.

The most amazing type of task to automate is expert tasks, which require deep knowledge and contextual understanding. Coding is the flagship example.

LLMs have consumed a lot of information and they are very good at small reasoning tasks. But, they still tend to fail on larger expert tasks. You can get around this by building an agent.

I want to stop here and define “agent.” An agent is a system that automates tasks through a combination of repeated LLM reasoning and pre-determined system logic. It is not just the large language model. Chat-bots are not agents in this sense unless they can reason about and perform actions on behalf of humans.

In order to take a complex task and build an agent a couple things are required:

• The task needs to be something that can be automated. Meaning it can be broken down into a series of steps that only operate on computer-controlled actions.

• The task to be automated needs to be well understood.

Turns out a lot of things fit the first condition. But, the second condition is a little more interesting and I want to explore it.

To build an agent you are assembling a system to reason about the task and perform necessary actions. This system needs to make many calls to the LLM to perform multiple reasoning steps, it needs to interact with other parts of the system, manage context, and it needs its own expert orchestration for the task.

Expert orchestration requires understanding the task.

Only an expert can provide this. But it also requires that understanding to be systematized. That requires an expert who can systematize the task, a separate ability.

The ability to systematize a task requires reflection and articulation of knowledge. You not only need to know how to perform the task. You need to know how to explain it. Because, well, you need to explain it to the agent in prompts and codify it in a programming language.

A lot of experts know how to perform tasks but lack the self-understanding to fully explain it. It lives mostly as tacit knowledge from experience. They can explain parts of their knowledge but that doesn’t mean you can follow their steps and perform the task like an expert.

I’ve seen attempts at building agents that fail at some part of this. Either an expert tries to build it but doesn’t quite understand the technical approaches for managing agents. Maybe they can’t quite explain the task accurately. Or, a developer tries to automate a task they don’t fully understand. All lead to lackluster, faulty systems.

So, building an agent requires:

• The technical knowledge of the abilities and limitations of LLMs and agentic systems.

• Expertise in the task to be automated.

• The ability to articulate and systemize the expertise.

It’s no wonder that coding agents have come so far. AI programmers naturally have the first two. If they have the third, they can build coding agents.

Final thoughts.

You can build a system that learns and evolve the expertise. This isn’t a bad idea but will start off inept. It happened with coding agents even though experts are building them. Progressive improvement is always a part of software development.

On top of that, LLMs have a second advantage at coding because coding is inherently a language task. But, LLM improvements alone are not the only driver of coding agent improvements. A lot of the agent quality depends on the agent logic which is why two running the same model can differ so much.

For an agent to start strong, you need an expert in the task who can articulate it clearly. And you need to pair them with a developer that knows how to build agents.

Coding agents are automating the second part. So, we’re going to see more and more expert agents.

There’s a common saying that naming is the hardest problem in computer science. Based on my experience this is pretty true. But it’s a hard problem for more than just software. Other fields like marketing and design need to name things too. It’s not much easier there.

The power of names.

I’m a pretty big fantasy reader. There this common theme where names have power. Knowing someones true name may give you control over them or remove their power over you. Knowing a true name of an element or natural force gives you the ability to command it.

These concepts are ancient. The concept that names can have magical or spiritual power dates back thousands of years. And there’s no coincidence here.

Names do have power. When you name a product, a feature, a concept, you shape its identity. Until you produce a name, the thing remains formless, almost ephemeral.

This is probably why people use meaningless code names like “Fudge Sundae” for products and versions. They can decide what the thing will be as they build it. But then, it also foregoes the power of the name for aligning understanding.

I personally struggle to work on an unnamed project. I need that clear reference and guidepost that comes from the name. It can change later if the project evolves, just like a Pokemon. No one wants Unnamed Tiny Sea Dragon. You want Dratini even if it eventually becomes Dragonite.

What’s in a name?

Naming as a skill sits behind the ability to articulate a concept clearly. In this sense, articulation is the ability to express a clear definition and then naming is the ability to condense it.

You can’t condense unless you know what you’re condensing.

To do this well we have to consider the purpose of and audience for the name. Names can be either a clear label or generative. A great name may be both.

Let’s consider the ability to collect things. We could call it

• Like

• Favorite

• Save

• Pin

and other names too.

“Like” and “Favorite” suggest they may also give feedback. “Like” is more feedback than save. “Favorite” is more feedback plus save. “Save” has connotations that it’s more private or even becomes local. “Pin” has connotations that it’s more public, shared, or involves some placement.

Different domains have different requirements for names. Marketing is going to lean more towards generative, evocative names. Software is going to lean more towards clear labels. But then there’s middle ground like product development. Products need to be clear and evocative which makes naming very hard.

This same nuance that shows up in saving features shows up in all the concepts in a product and the interaction between them can become eye-watering. I’ve seen systems that have several different saving concepts and now your users and your team need to know the difference.

Crafting names.

There’s a lot that can be said about creating names. Whole books have been written on it. It’s incredibly nuanced and contextual.

When a name is hard to find people often jump to just listing words or phrases. What ends up happening is it stalls out and a placeholder or subpar name will get chosen.

My advice here is to step back and find a definition before trying to find a name. Again, you need something to distill. But then, the best name may not be a direct distillation. It could be something that evokes the definition or a feeling we’re trying to create. You need to know what you want the name to do.

Finding a good name requires finding a word or short phrase that is accurate and distinctive. It has to be a clear reference without accidentally becoming confused with something else. It needs to achieve the goal for it’s audience. It’s a commitment and bet on an unclear future.

All of these needs to happen at the same time. This is why naming is the hardest problem.

When developing products I receive a lot of feedback. It’s a good thing but the form of the feedback is often lacking. Then you have to sort of divine the actual meaning behind it.

It’s not a big deal when this comes from users. It’s unfair to expect users of your application to give extremely clear, insightful criticism. Feedback from users is more about knowing there’s a problem and ideally where it is. Anything beyond that is a bonus.

Don’t give descriptions.

People who work on products for a living struggle to give clear feedback just like users. At the worst end you’ll get vague descriptions and feelings that are barely even developed.

“Hey, this page looks a little weird.”

A little more useful, sometimes you’ll get specific problem statements that are still too vague to be actionable.

“We aren’t getting clicks on this page.”

“These buttons on the home page look off.”

The problem here is that this leaves the person receiving the criticism to identify the root cause and what the shape of a solution should look like. Maybe that’s their job but quite often I’ve seen low quality criticism come from the core stakeholders and decision-makers. Then you end up playing an expensive game of product Pictionary where the guesser knows the answer and won’t tell you. So you keep redrawing until they are satisfied and say, “yes that.”

Make claims.

When delivering criticism internally, it should come with a claim. You need to state a problem and give a reason for that problem. Even if it’s only a hypothesis.

“People aren’t clicking anything on this page because the copy is unclear and the button blends in.”

That is testable and actionable. The claim may end up being wrong after testing but at least we’re now on a path to resolving it.

Forcing the “because” to be a part of the criticism is an easy heuristic for improving the quality of that criticism. Other improvements could be setting expectations and giving context. These are things I often need to fish for after hearing vague descriptive commentary related to a problem.

“We expected people to see this and sign up. Research suggests 5-10% of users should take action here but only 1% are. I think it’s because the copy is unclear and the button blends in. But maybe we’re driving the wrong people to the site or this isn’t compelling enough.”

Now this can become a product discussion with many actionable outcomes.

Software projects have this habit of becoming overbuilt. Feature creep comes from various common pressures:

• We don’t know what our users want most? Give them everything.

• What if they run into this weird case? Build a feature for that.

• Oh! Dude! What about this cool feature idea!? Throw that in too.

Somehow, a lot of people working in software don’t really see the problem here. At the very least, they don’t take it seriously. Building like this creates a graveyard of unused features and wasted time, sure.

So, what’s the actual harm in that? This is one of those problems that’s hard to model and quantify.

1. Overbuilding wastes your time.

This is the obvious one and the easiest to put numbers to. When people call out feature creep wasting time is usually the argument that goes with it. You build things no one ever users just because someone might want them. Or, in my experience, quite often because someone internally wants it but customers do not.

It’s not just a one time cost either. As you add more and more features and paths to your product the cost grows exponentially. Every flow requires test coverage. There is more code to manage. Product guidance and marketing need to be kept up-to-date.

If this problem is hard to model, it’s only because the complexity of the downstream costs are so enormous. It’s easy to focus just on build time and forget it’s only a fraction of the actual work created from a code change.

2. Overbuilding creates and indicates organizational problems.

I don’t hear people bemoan this one that often. It stems from the downstream costs of wasting the time of everyone on your team. You end up having meetings and arguments about features no one will ever use.

Feature creep creates more things to measure for success. Then people need to investigate why they are failing. It takes team focus away from more important product decisions. As some features are failing it dilutes clarity on the overall product success.

Before that even happens, feature creep a huge red flag for the product. It means you either don’t know what to build or you’re scared that your product thesis is wrong. So, you’re either shooting in the dark or hedging your bets.

What’s worse is that feature creep can feel like productivity while masking these issues.

3. Overbuilding waters down your product.

Then there are your poor, unfortunate users. If you’re confused about your product, just imagine how they feel.

Your marketing probably doesn’t convey a clear message because your product doesn’t. People will go in expecting a specific solution and struggle to find it. Or, feel that it’s overcomplicated.

You may have a few very useful features in your product. It’s too bad people have to click through 30 different things to find them. Or make 10 decisions when they really just wanted a one-click solution. Instead of having a focused easy-to-use product you have a frustrating, confusing experience.

Operationalizing “no.”

Awareness of the problems with feature creep is a good start. Yet, I’ve been on teams that were aware they were doing this and kept on doing it anyway.

You need to have a culture of “nope.” But to do that, you need to understand what you’re building and why. If you don’t have clear core flows in your product, it’s all too easy to just keep adding more. You throw shit against the wall to see what sticks. Eventually the wall crumbles.

It’s okay to generate and log ideas for building on the core product. But until the core product is working and people are asking for more, there is no use in building more. Save those ideas for later. Keep them in mind and build extensibly. They are immediately useful for designing forward-thinking systems for later expansion.

Ruthlessly prioritize your core flow. Anything that doesn’t make this flow clearer does not belong here. This will require some process work. It’s easy to say yes and throw some idea on the list. Usually saying no is more work. People will defend feature ideas and generate them constantly. There needs to be a systematic approach to defending the priorities.

Only once you have feedback about the success of your product’s heart can you start adapting and adjusting it. Doing so beforehand is just letting fear and chaos reign.

When I was in college I studied Economics. If I could go back in time I’d kick myself but that’s the choice I made.

While studying I observed this tendency in modern economic theory towards looking at what happened and building a mathematical model to explain it. From this model you would try to extract predictive power.

Further, all of these models assume “rational behavior.” I’ve never once in my life seen rational behavior.

None of this ever sat well with me.

If these models had extreme predictive power we would be able to reliably predict a lot of major economic events. Read the news and you’ll see that’s hardly the case. Then there’s the fact that these models are post-rationalizations. Sure they try to cross-check under different conditions but economics isn’t really an experimental science.

There’s a saying I learned while studying that goes “ask 100 economists what they think and you’ll get 101 different answers.” The implication is that economists basically all operate on their own individual viewpoints. And even then they can’t always agree with themselves.

The main thing my study of Economics left me with was a tremendous amount of doubt in statistical models. Not what I was supposed to walk away with but it was probably worth the price of admission.

The weird thing about the world is that there is very little we can actually know for certain. We build mathematical and rational models and use them to predict the world. Some of them are very, very good models. But they’re still just models.

People often think they know things, but that’s different. People definitely want to believe we can know things and reliably predict events. Certainty makes us feel nice and safe. It helps us sleep at night.

This desire can be a little dangerous. Trusting dodgy models creates overconfidence. In fact, the definition of “overconfidence” is exactly this. You have mental models of how the world works that you think are better than they are and you trust them too much.

Then there’s this other tendency that pairs up nicely. People would rather go with a plan that relies on a model than one that goes on any other seemingly irrational, unquantifiable approach. If you can make a statistical model and compose a rational sounding theory, people trust it much more than an idea that is based more on experience and heuristics.

This kind of thinking makes sense on paper of course. But it discounts experience and forgets that models are a type of heuristic.

Let’s try to make this more tangible.

One place that I’ve often seen modeling and reasoning win out over “I have a very bad feeling about this” is outsourcing. I’m a software developer and I’ve been on many projects where people try to outsource to cheaper developers overseas. This makes a lot of sense on paper and you can see all the companies out there doing it and feel safe.

Several projects I’ve been on I was actually brought in because this went less than well.

Here’s the problem with outsourcing that quite often happens. The people you are outsourcing to are paid hourly and not much at that. They are in a wildly different time zone. They don’t actually care about your product. They want to work as many hours as possible to make as much as possible for as long as possible. They are going to do the bare minimum to keep the project going but aren’t going to try to make your product better. They’ll just do what they’re told in a half-assed way and that’s that.

If you’ve worked with a few outsourced projects you likely know what I mean.

The problems here are very hard to model in numbers. I can make a somewhat rational argument for why it’s a problem but I can’t easily assign dollars to it. Some projects seem to do okay with outsourced development. Many stagnate and fail entirely.

Models aren’t inherently bad. They give us insights and help understand the world. But we can't fully understand the world. A fact worth remembering. Models are useful for guiding us towards bets and experiments but practical experience and real-world results are really what matters.

Always question models because reality has a way of doing something else.

As the attention economy has evolved, social media has become much more personalized. Optimizations in algorithms keep you attentive and engaged. They lift things from the ever-growing pool of available content to put them in front of you. All of this exists to keep you watching or scrolling. These algorithms learn with ruthless efficiency.

Social media has a limit, though. Someone has to create the content for you to consume.

AI does not share this limit. It can generate whatever a system decides you want to see.

ChatGPT was the fastest growing product of all time. It reached 100 million users within 2 months of launch. For comparison, TikTok took nine months to reach this user count.

LLM apps like ChatGPT and Claude let you ask for nearly anything you want. They can research esoteric topics and talk about your problems with a system that is always available. At the same time, we have rapid evolution of image, video, and audio generation models.

All of these together enable the creation of media that is just for you. This is coming at a time where media consumption only continues to rise.

I find this new reality both exciting and terrifying.

It’s exciting because it empowers learning and building rapidly. I run research queries regularly and answer questions might have taken days or weeks in the past. I’m not limited by existing research papers or the time it takes to source them.

Right along with that, I can build applications with unbelievable speed. This lets me try things out without spending days or weeks. In many cases, I can test out a feature idea within hours. Then there’s all the fun things image models can do.

It’s terrifying because it has the potential to amplify all the problems of the attention economy.

I worry about the societal costs of constant media consumption. There’re the echo chambers, the seemingly correlated rise in mental health issues, constant notifications, and just how much of your life it takes from you. Then you multiply that across the entire population.

The jury is still out on the impact of AI here.

My biggest problem with social media is that it’s a random reward system which can be highly addictive. Personal media, driven by AI, and requested by the user doesn’t have this random quality. So, you can get what you want and move on with your life.

I’m holding onto some optimism that this could lead to a turning point in our relationship with media consumption.

Then again, I don’t think personal media will really replace social media. Not in the short run, anyway. The much more likely future is that social media continues to grow and consume our time. Meanwhile, AI will speed up our other work and give us more time to consume.

It’s not the the future I want to see, but it’s the one I worry we’re headed towards.

When you use a product there are a lot of perceptions during your experience. Some of these are strong feelings, both positive and negative. Most of them are subtle observations that people would rarely every call out.

Let’s say you’re using an app and there’s this small, almost imperceptible hang at the end of every interaction. You subconsciously notice it but don’t think much of it. Let’s say the same app has a slightly dated UI. Again, you don’t think much of that. But a series of small feelings like this can build into a lackluster feeling about the app. It can feel worse than its pure utility.

The opposite is also true. Maybe you buy a coffee machine that has really nice detailing and finishes. It’s missing a few bells and whistles but you feel like it slightly elevates the beauty of your kitchen. In this case it feels better than it is.

The truth that a lot of product teams seem to ignore is that aesthetics and vibe are an important part of a product. We all want to use products that make us feel good.

Product development teams will hand waive this part away. They think of it as noise at worst and nice-to-have at best. However, vibes drive the disposition people have towards your product, especially over the long run.

Most teams start products from a stance of “make it work, then make it good.” We iterate along and get to a point where it’s ready to release. But in most product cycles you will stop well before good and certainly before feeling good.

There are several drivers that downplay vibes from the engineering side of product development.

• It requires operationalized taste to be cohesive.

• It’s hard to rationalize into value.

• It’s not something you can measure.

The lack of operationalized taste is the blocker I see most often. I’ve been on projects that value “look and feel” on paper but don’t manage to execute on it. There’s no one on the team who actually owns it. And, even if there is someone driving it, they often get overruled by other stakeholders.

To make it work, the owner needs the ability to say “this doesn’t feel right” even though it works. And that has to matter.

Their taste is a dowsing rod for the product’s feel. This needs to get codified into actionable decision heuristics and principles. Taste isn’t something you can create deterministic rules about. Rules may be more like prioritize play in these situations and clarity in these.

All design and copy changes need to go through review. These types of changes can be automatically detected in software pipelines and the PR can be flagged.

Realistically, this is a hard part of product development to fully operationalize. Only one company I’ve worked for bothered with anything like it. Design quality was a core part of their brand.

Organizations that truly value the form and feel of their products are rare. It takes more effort and vision. When they show up, people notice.

Just remember people also notice when you don’t value your product’s vibe.

I was just talking to my wife and I realized it’s getting late in the day. So I said, “Welp, time to go write my daily shit stack.”

At the end of last month I decided to write on Substack every day this month. I plan to stick it out to the end because I tend to be a man of my word.

But I have some regrets.

Publishing daily is a pain in the ass. I quickly ran through the obvious writing topics bouncing around in my head. Each piece feels a little half baked. Now, every day I have to really dredge my brain for something to write about. I expect future writing will be a little less frequent and more developed.

So, why did I decide to do this?

My main goal is to get some ideas out there and get better at writing by doing it more often. That’s the only way you can.

Of course there’s a deeper motivation.

I work at a startup that is trying to build the future of job searching. We have extensive professional profiles and job matching and loads of data.

During this project I’ve tested our own application and a lot of other applications that help you find a job. While trying these in earnest I kept running into this constant issue. All of the suggested jobs on all of these platforms suck.

How can job platforms all be so bad?

There are some very real answers to that question. Maybe another time. For myself, though, I realized I just don’t like working on other people’s ideas. It’s brought me constant stress throughout my career.

So, I decided I need to do something about it.

I don’t foresee a near-term future where Substack pays my bills. I don’t even see a short term future where anything other than a regular job pays my bills. But, it’s time to start taking steps in that direction.

It’s hard to make major life and career changes in a single motion. This is something else I’ve learned on this project.

Career changes are usually small and directional.

In Directional Living, Megan Hellerer describes how she had a similar realization about not enjoying the career she had built. She took a drastic step and left it suddenly. However, afterward, she had to take more directional steps to find the right thing and eventually wrote her book.

Directionally, I’m looking to explore my own ideas. Making them public is an easy first step. Unlike Megan, I do enjoy what I do. I won’t quit suddenly. I like building things and trying them out.

I don’t like building things I expect to fail. That’s something I’ve had to do many times in my software career.

It feels like a waste of my precious life hours and you only get so many.

For a long time taking notes was my main method of gaining value from books and other media. I’d pull quotes and note interesting concepts with references back to the source. For a while I’d make notes in the margins or sticky notes attached to pages.

That’s what I was taught that taking notes should be.

It feels productive but isn’t really. It’s a form of intellectual hoarding.

I’ve been getting rid of older books and I see these all these margin and sticky notes. I find note cards and other artifacts. And yet, I doubt I could tell you about the content of the books. I suppose I could flip through and look at the notes but I don’t see what there is to gain.

Unless you put ideas into practice, there is no value in collecting them. Hoarded ideas only fade away.

I’ve become a lot more selective with the books I read and the way that I engage with their ideas. Some of these books no doubt had ideas that would still be valuable. There are a few that I plan to read again with a different approach.

Now when I read a book I will stop when I encounter an idea that is actually thought provoking. If the ideas in a book do not cross this threshold they aren’t really worth my time. After stopping I’ll write out a few paragraphs that integrate the thoughts with others that are relevant to me. If I can’t do this, then it wasn’t really that thought provoking.

From the outside this can look like the same thing. You read and then you write some stuff down. But capturing unrefined notes is just a collection habit. Thinking through writing is synthesis by externalization.

Quite often these are ideas relevant to my work or personal projects. So, I’ll go further and fire off an email or create a document in the project that further expands on the idea.

As a result of this change, I read less and more slowly. Sometimes it takes me a month to get through a good book. The more useful the book, the longer it takes me to get through it.

I also feel no remorse in abandoning a book that isn’t providing useful insights. Completing a book isn’t a reward in itself. It’s silly to force yourself to through just to gain nothing.

Productivity media tries to tell us we should be trying to consume more and more information. People challenge themselves to read dozens or hundreds of books per year. I can’t imagine getting real value out of a book read that fast. It’s just performative and self-congratulatory.

All of this applies to other media as well. I’ve limited the amount of YouTube and podcasts I listen to. I got tired of hearing the same things and doing nothing with it.

There are better uses of my time. I bet the same goes for you.

I frequently feel like an asshole.

I’m hypercritical. I have strong opinions about the products I work on and the ones I use. These opinions come from years of evaluating products for why they succeed and fail.

My career started in QA where I discovered a knack for breaking things, bringing up weird edge cases, and pointing out inconsistent design. Since then I’ve moved into a hybrid role of development and design and my ability to identify flaws in design, architecture, or even strategy has only sharpened.

My biggest issue with being in QA was that I constantly felt like a teacher marking red ink all over other people’s work. This feeling was my primary driver to move on to development.

You might think this means that the things I create are near perfection. But, that is very much not the case.

The reality is that building things is an iterative process. We have to design and build and then find flaws with what we have. This brings us to another round of designing and building. In modern software there is never a done state.

The ability to rapidly find flaws is truly a gift. The sooner they are brought up the faster the iterative process of improvement goes. The downside is this skill is often perceived as negativity.

People I work with know me as hypercritical. Some see it as great taste and some see it as being a nay-sayer.

A lot of this comes down to the culture of the team. Some teams are more open to criticism than others. I’ve worked on projects where whatever leadership says is what gets built. Those projects are all dead.

People often tie the products they’re selling to their egos. It’s a dangerous situation because it makes you willfully ignorant to the product’s problems.

I have to choose battles carefully because sometimes I need to bring up critical product flaws. These need to be clearly separate from minor design and brand risks.

All criticism can feel like an attack. A deluge of criticism feels like a war.

A hypercritical person needs to be careful with how they present their thoughts. If you use vague language and derogative phrasing your criticism is going to raise defenses instead of consideration. I find it’s often best to open these discussions with questions rather than statements.

This is a part of the skill that can be learned. Observing reactions to criticism is a wonderful teacher about approach. When you see guards go up, back off and try a different tact. I often see people try to deliver criticism without specificity or context and it’s not productive.

The eye for problems is harder to learn. There’s a required combination of system knowledge and context awareness. You need curiosity to prod things from every angle.

Curiosity for problems is more a thing you develop and less something you teach.

This is why some people see criticism as negativity. It’s an approach to looking at a product that not everyone engages with. To avoid being an asshole, critical people need empathy for others that don’t see this way.

Resumes are universally written as a list of timespans. We list out the time we spent at this and that company. We describe the things we did there. Maybe we list some keywords for ATS systems to pick up. If we had a significant deliverable or outcome it’ll get a nice little bullet point.

I’ve looked at hundreds, if not thousands, of resumes and they are all boring. Mine is boring. Yours is boring.

There are some things they can tell us about a person, though.

I’ve landed interviews partly because of the paper I printed my resume on. Some years ago when I still had to drive around and hand out resumes I opted for higher quality paper that cost a little more. I’ve had multiple interviews start with a comment on the quality of the paper.

Learning from this, a big thing I look for in a resume is whether the person has attention to detail. Is this well formatted? Are there typos? Is it consistent and aesthetically pleasing?

These come from the design of the document itself and not the content. It’s also a problem largely solved by ample resume templates and AI generators. Resume tools will now rewrite a beautiful resume for every job you apply to. It makes the document itself largely useless.

The utility is slipping away. I can no longer tell if you opt for the nice paper or go for the cheap and thin stuff.

A career is filled with moments that define a change in direction. These could be a sudden realization, a change of worldview, or shift in priorities. There’s a story behind that change. Why did you do that? What did the experience teach you? Why did you leave this job?

It’s absurd that we have to wait until the interview to dig into the interesting points in a career. Yet, resumes are not designed to deliver this kind of information.

Instead, questions like these require context and narration to answer. A resume is designed to be something we can skim to filter people out. It’s an imprecise tool at best.

Arguably, if you maintain a history of publication it can capture these stories. But then, are we going to expect everyone to output content as a type of resume? It seems like a lot to ask. Yet, it feels like more and more knowledge work is moving this way. There’s this latent expectation to maintain a personal brand.

As I write this I wonder about creating a resume with a different kind of framing. What would it look like and how would people react to it? I imagine in the world of ATS auto-filtering, you’d be invariably rejected without a human ever seeing it. So this is mostly a thought experiment unless the system shifts to support it

In my current role we built a more narrative professional profile. It’s a step in this direction but still largely leans on traditional resume concepts. It bothers me.

There’s this underlying issue that the document still needs to serve as a filtering tool. The filters being uses are very shallow. They’re keywords and timeframes. But people aren’t checklists like this and transferable skills are very real. Most filters are wishful thinking.

When I interview someone, I’m really trying to understand their mindset. No document I’ve ever seen accurately captures this. I don’t really know how you’d filter on it.

A post-resume artifact will find a way to solve this problem. It’ll avoid keyword filtering as a core concept. It’ll help you more deeply understand the person behind the document.

This question gets asked a lot because people feel threatened by AI.

There are days where I wonder about the future of my career due to AI. I’m a principal level engineer and I’ve fully converted to agentic development. Anthropic recently released a labor market impact report that suggests wide impacts to knowledge workers that are yet to be fully realized.

There are two compelling answers that often get floated:

1. Humans can decide what to do.

2. Humans have taste.

To me, these are the same answer. People often think of taste as this sense of good and bad. But really it’s about making the right decisions for the given context. It’s deciding what to do.

But, why can’t AI do this?

Agents make many decisions given the context of a prompt. It has taste. I’ve seen it take a single prompt and make reasonable applications. I’ve seen it make beautiful user interfaces.

The element of decision making AI doesn’t have is initiative. AI requires a directive to do anything. AI isn’t actually embedded in the world in any meaningful way. It can churn out ideas. But those ideas are not driven by lived context within the world. They certainly are coming solely from the AI.

Along with this AI has no continuity. I’ll be working with a coding agent for a while and suddenly it’ll forget critical details. This will get better but continuity is not just about memory. Every call to the LLM has to include context or it’s a fresh start. That context has to be managed. AI isn’t an entity embodied through time. We just use it to try to simulate that.

I see taste as a form of critical decision making. However, it goes beyond what AI is capable of currently. Human preferences are not just conceptual ideas embedded in language. Nor are they limited to sound and image. We are capable of processing experiences that combine all the sense and deciding if they feel right. Emotion is a critical aspect of taste that AI is not fully capable of appreciating.

I’ve mostly agreed with the common answers, though the reality is more specific than just taste and decision making. Human differentiation lives in nuanced aspects of those things, not the capabilities in themselves.

There are a couple human behaviors that AI is even weaker at.

1. Humans can be surprised and surprising.

2. Humans can be irrational.

LLMs are an averaging of many human outputs. They rarely give surprising answers. But humans can be very divergent from each other. And themselves over time.

You can introduce chaos factors into AI but there’s something different here. Humans can be surprising but internally consistent. When we try to get surprising answers from an LLM it tends to feel random.

Related to this is irrationality. The more knowledge used to train LLMs the more rational they tends to be. The ability to reason is essentially the goal of larger and larger models.

Sometimes we need something other than rationality.

Irrational answers are frowned upon. But, humans aren’t really rational. We have emotions and instincts. We need things to feel right instead of just being right. AI can rationalize an answer but isn’t equipped with the empathy required for correct irrational answers.

So then, what is the answer?

AI can simulate human thinking to a remarkable degree.

Yet, humans have to embody their decisions and the consequences. Humans grow and continue in a way that AI cannot. AI can mirror this but never truly be it.

We want a clear, certain answer to this question. We want to be able to say focus on this and your career is safe from AI.

Humans are fuzzy and so is AI. The answer has to be fuzzy too.

There’s a lot of overlap but taking the initiative to worry about a problem like this is something only humans can do.

There’s this famous anecdote from Art & Fear by David Bayles and Ted Orland. In the book he talks about a pottery class that split the students in half. Some focused on quality to make a single excellent piece. Others focused on quantity to make as many as possible. The students focused on quantity learned faster than the ones focused on quality.

Even though this is an anecdote, research supports the idea. There are research studies looking at quantity versus quality in learning and creativity (see list below).

The general consensus is that quantity outperforms.

Quantity yields greater diversity of output and promotes divergent thinking. From a pure chance view, you’re more likely to create something great if you create more things.

Quantity gives you more fresh attempts to learn from. It creates fast feedback loops and gives you chances to compare and contrast. You get more experiments in.

More anecdotally there are design methods that rely on this reality. There’s the Crazy Eights and thumbnail sketch exercises try to cover more ideas. Parallel prototyping and A/B testing evaluate multiple solutions to a problem.

Quantity is effective for generating ideas and learning quickly.

People, especially perfectionists, will get hung up on needing to research and study before starting something new. But, the best answer is to just start and do a lot of it. Don’t worry about messing up and failing because you need those experiences.

I’ve unstuck myself this way a few times.

Some years ago I wanted to learn to paint. I found myself watching videos and reading books on how to get started. Eventually I said, “fuck it” and decided to just do at least one drawing or painting every day until I hit 1,000.

I’d just read Art & Fear.

Anyway, I tried a bunch of approaches and techniques. I worked through some courses but always hit the very low bar I set for myself. Some of the things I did were total garbage. Probably most of it.

But in the end I started doing plein air style paintings from video games and ended up being part of an art exhibit on the subject.

Currently I’m doing a daily article on Substack because I wanted to get better at writing. The goal isn’t as aggressive, only for the month of May. This is the fifteenth article.

Going forward I’m planning to do one of these monthly for different outputs.

Reframing a goal to be about quantity instead of quality changes the way you approach it.

In The Score, C. Thi Nguyen describes how collapsing goals to numbers can lead to value capture. That is, instead of focusing on the values behind the goal, you start valuing whatever hits the target. Those values may be misaligned. Much of the book is about how this is a problem.

That misalignment is a feature here because our values weren’t quite right. If we overvalue quality we get stuck just trying to understand what that means. And in reality quality is subjective. It’s hard to even understand it without an interactive feedback loop of experimentation.

Setting numerical goals for output makes them easy to gamify.

Nguyen also discusses how games use numerical goals but they feel different from metrics. Games let you try values you may not otherwise have. They can also help you engage in “striving play.” Instead of trying to win like an “achievement player,” a striving player adopts the rules of the game to enjoy the struggle, challenge, and joy of the pursuit.

By gamifying your learning goals through numerical targets you can build an appreciation for the feel of the endeavor itself.

For example, let’s say you want to learn to paint. You can do 5 paintings like Bob Ross. You can do 10 like Andrew Wyeth or J. M. W. Turner. You can try painting scenes from 20 games. I ended up setting challenges like this and learning a lot from each one.

These challenges help you learn directionally. You won’t go from amateur to master through a single set. You won’t find your own unique voice or style immediately. But by trying different things you find what you like and don’t like. You build your taste and your skills.

This is the part that truly matters.

Your vision of perfection will shift through time but your knowledge only grows through experience.

Research papers on quantity over quality:

• The iterative and improvisational nature of the creative process

• Measuring and Motivating Quantity, Creativity, or Both

• Quantity and/or Quality? The Importance of Publishing Many Papers

• Beyond Productivity Loss in Brainstorming Groups: The

  Evolution of a Question

• Quantity yields quality when it comes to creativity: a brain and behavioral test of the equal-odds rule

When I was in high school and college I worked for Domino’s Pizza. It was a good first job but there was one thing in particular that I loathed. You would be scheduled for a shift starting at a specific time but you did not know when your shift ended.

It worked this way because management could not know how busy it would be. Once business started to slow the shift manager would start cutting staff for the night.

This was the absolute worst on the weekends. Friends would call me while I was working and ask when I was getting off. I could never tell them and it always annoyed me.

There’s something about uncertainty that is just absolute torment.

A couple years ago I was working for a startup that suddenly went belly up. The details don’t matter but we were expecting a new round of investment and suddenly they pulled out when were were already stretched thin. Everyone was let go that day.

I spent the next few months unemployed and the torment of uncertainty reared it’s head again.

Job searching has never been a pleasant experience. However, the current job market feels extremely rocky. You have more information than ever. More jobs to apply to than ever. It’s easier than ever to apply.

But, then, that access and ease is the root of a big problem.

Applicants are overwhelmed with options. You can apply to a hundred jobs per day with Easy Apply or AI resume slopguns. But then, employers are on the receiving end of that. They get swamped and they get overwhelmed. They don’t reply to everyone because they just can’t.

Job seekers are in uncertain waters because they don’t know how they stack against the other 5000 applicants. They don’t know if they’ll hear back. So, they turn to bigger and better application padding.

Employers don’t know how to deal with what they’re getting so they adopt questionable AI tools for filtering and ranking. People get sued. People get apprehensive.

In situations like this we can’t always fix the root issue. People demand easy application tools. People love remote work and options. Employers benefit from options too, even though it’s overwhelming. Unless we just let the AI overlords assign jobs there’s no easy solution.

But, we don’t need absolute certainty. We just need a little.

I’ve been reading Alchemy by Rory Sutherland. In the book he talks about this issue and gives a great example. Would you rather:

• Your flight is delayed.

• Your flight is delayed by 40 minutes.

Most sane people will pick the latter. At least you know what to expect.

We can’t fix all the problems in the job market. We could do a better job of tracking the process.

An employer doesn’t need to respond to 5,000 people. They just need a tool to say, “Hey we’re interviewing a few folks now. If it goes well the job will be closed. If not we’ll interview some more.” Like a Pending status on the sale of a house. And then, the applicants just need a way to see it.

We can create tools to help people understand how likely they are to get a given job. Similar tools can be built to inform the employer about the status of applicants.

Problems like this don’t need to be over-solved. We just need improvements that consider the human struggle.

Feature creep creates a lot of problems. The usual suspects are real: lengthening time to release, overstretching the team, technical debt, poor user experiences.

The worst thing that feature creep does is blind you to deep product issues.

Let’s say you’re building some new app and keep packing more and more features in there. You build whole systems and infrastructure. The UI explodes in complexity.

Now, you finally release it. You start hearing about loads of issues. This page doesn’t look right on mobile. This widget doesn’t save correctly. This image doesn’t look right when it loads. The list goes on and on.

So, you start fixing all that. It’s an all-hands-on-deck emergency.

But, you’re ignoring the most critical questions after a new release. Does the app solve a valuable problem? Do people even want this thing? Are people sticking around?

Let’s consider the less likely path. You release and collect user feedback despite all the issues. You get metrics in and review them. You try to answer the core product-market fit questions.

Your attempts will be muddled.

You aren’t the only one blinded by feature overload. Your users are too. As you collect feedback user aren’t talking about whether it solves some problem. They’re telling you about bugs and strange layouts.

The only way to win this game is not to play.

When I’ve seen this happen, it’s always because the product thesis isn’t clear. We know the general shape of the problem we’re trying to solve. Let’s figure it out as we go. So, you build Swiss army knife instead of a corkscrew. Sure there’s a corkscrew in there, but it’s a crappy one.

You shouldn’t be building anything until you know exactly what problem you’re trying to solve or how you plan to evaluate your solution.