Journal

I fell for a trap I never saw coming, and even when it arrived, it took me a while to notice.

Cabin's initial design

I initially designed Cabin to not be a basic grid of cards. After thinking for a bit, I settled on the following:

1. One big entry card, featuring the latest addition;
2. One horizontal row of recent additions excluding the featured one;
3. And an infinite-scroll grid of the rest of the content.

I also chose to not have any subjective ranking among the included videos, so I left the chronological order as a default.

That became a huge issue.

What broke down, and why

A few, silent psychological issues sneaked in.

The featured card always was the last video and it set the tone for the whole website. That added a silent pressure on me to reach out for more "prestigious" topics - something like science explainers. I had trouble choosing anything else - "this sets a wrong tone for the website. Let me find something that sets the correct one". But that's a wrong question - the layout shouldn't be dictating what I include.

Further: Since the sort order was chronological, it basically followed my watch order, and topics clustered. This created a pretty big problem - Cabin started to develop an identity, accidentally. It started to filter for topics, and became like a science explainers hub. That was NOT a goal. Cabin is supposed to contain ALL kinds of videos, as long as they meet the quality bar.

So, what am I left with? I include mostly science explainers, they are clustered by my watch order, and Cabin has an identity it didn't mean to have. All of a sudden I'm stuck on 3 channels, 1 topic, and can't add anything.

Cabin, and me, are stuck.

The redesign

Took me quite some time to realize - watched some really good videos, still wasn't sure if I should include: "should this topic be included?". THAT SHOULD NOT EVEN BE A QUESTION. If it does its job well, then it should be included. Doesn't matter if I'm interested in the topic or not.

So I observed what was the blocker, and I arrived at the signal. The featured card, and the topic clustering by chronological order would greet anyone with the latest videos I watched and included, which would instantly set the tone for the website. The problem wasn't that I included those topics, the problem was the signal it sent to a newcomer. So I broke down the signal.

Now Cabin is randomly shuffled with an ability to sort by the newest or oldest. There are no featured cards, and every entry is uniform - same size, same card, randomly shuffled per visitor per session. Now even I have no idea what's gonna be shown up at the top, and how the topics will be clustered.

Now the real problem showed - Cabin is too narrow, mostly science explainers. This was the signal for me, the one I wanted - now I'm pushed to discover other topics, to fill the gaps, and bring in a huge variety of interesting topics and videos.

Now I believe Cabin is set up technically, and the only remaining problems are filling the catalog, and marketing..

This is the first actual journal note I'm writing and I have no clue what I'm doing.

And I especially have no clue what I should be doing, and am not.

Trying to discover good content has become DIFFICULT

Doesn't make sense, right?

It does.

The world became too addicted to clickbait-y titles, topics, and framings. They get most of the clicks, so everyone started optimizing towards it. This caused genuinely good creators to start blending in - morphing their titles and thumbnails into more generic, clickable versions.

But why would they?

Because retention determines who gets popular, but your video has to be clicked in the first place. Titles became too similar, so distinguishing between great and slop content via titles and thumbnails semi-reliably isn't all that easy anymore. Everything gets clicked, including slop, and it's in a VAST quantity, drowning plenty of talented creators in rankings by the sheer volume.

Which means, when I'm trying to discover new creators, I'm struggling to find unknown ones - big names show up, slop shows up, relatively lower quality shows up, but relatively unknown creators don't.

Which is exactly the problem I'm trying to resolve - making high quality creators and content easy and straightforward to discover - now facing it myself, firsthand.

Why this is a pretty big problem

The entire premise of the test run is to check whether people actually care about having a place where everything you click is of a high quality, guaranteed - therefore never having to worry about clicking anything and leaving dissatisfied. More precisely, minimizing the chances of that.

Without sustained discovery of top tier content in a decent volume there are no new entries to add. Without new entries people don't come and Cabin doesn't grow its own audience. Without audience, there is no recommendation of such unknown creators. Self-closed loop - the chicken and egg problem.

Without audience, there is no reputation. Without reputation, no one takes Cabin seriously, therefore, people just scroll through Cabin's call for such creators.

How I'm fighting this off for now

Personal outreach to people I know. Asking for recommendations for relatively "smaller"-sized channels, something like <150K subs, and going through their content.

This works, but it's not a sustained volume. And I encountered the issue I described above - those creators trying to blend in, morphing their title, and causing mismatch in the premise and the delivery. Even tho their content is really high quality, it isn't what I expected from their title. That's disappointing.

More on that later, maybe.

Wanna help? recommend creators/videos via Cabin, or even better - on X or Bluesky, so Cabin gets more reach and more people see my work.

Ungabunga. Ungabunga bunga, unga. I want to see what a paragraph looks like when it is long enough to wrap onto a second line and then keep going for a bit after that, so here is some more of it, wobbling along.

Wobble the first

Snorf. Grimble grumble bold wumpus and then slanted noodle and then a bit of inline code smell right in the middle of nothing.

• gronk
• gronk again but more so
• the third gronk, which is admittedly derivative

A smaller shout

Plimf. Plimf plimf.

Behold, the quoted nonsense, which is indented and greyer than its neighbours and spans two lines to prove it joins up properly.

1. first thing
2. second thing
3. third thing, arguably the best thing

Bimble. I am checking that a link to nowhere useful still looks like a link, and that an outbound one picks up its own treatment.

Wobble the second

Florp florp florp. Nothing here means anything and that is the entire point of it.

Wibble.

That is genuinely all of it. I want to see whether a two-line post looks broken in the listing or just looks short.

Hark. The snorkel committee convened at dawn and immediately fell over.

What was decided

Nothing. Gronk was tabled. Plimf was struck from the record for reasons nobody wrote down, which in hindsight seems like the kind of thing somebody should have written down.

I am mostly testing whether three posts stack sensibly and whether a summary this long wraps without looking like a mistake. Blorp.