There's no such thing as a good film.

(Same goes for series, documentaries, anime. I'll say "film" to keep it short.)

For years, nine out of ten films I was told to watch were wrong for me. Not bad. Wrong.

If you've ever lost an evening to scrolling, stay. This one has a happy ending.

It took me years, one AI, and a hundred blunt little notes to get there.

The odds were never on my side

A top 50, a trending row, a critic's year-end list: they're built to please as many people as possible. It's a numbers game. And nobody's taste is exactly the average. Mine was just further from it than most, so the numbers were against me more often.

If you've ever given up on three films in one evening and put The Lord of the Rings on again, for the hundredth time, you know exactly what I mean. (For me it's that, Louie, or The Matrix.)

So I fought them. I learned which sources to trust, which keywords to type, how to drop a title after two lines of synopsis. It took years. It worked, more or less.

More or less is a terrible way to spend your evenings.

Most recommendations are built for the middle

"Best film of the year." "Everyone's talking about it." "96% of critics agree." None of that is about you.

The studios and the platforms aren't evil. They just make more money when we all want the same thing. One hit for a hundred million people is cheaper than a hundred million different evenings. So everything pushes you toward the middle, and calls it taste.

The more alike we are, the easier we are to sell to.

Don't get me wrong: I love plenty of big hits. House of the Dragon got a 10 from me. "GOT and company are the peak of entertainment." Interstellar, Gladiator, Dune: all favourites. Liking what everyone likes is fine.

The point is to know why you like it. To love a film because it's good for you, not because a banner told you everyone else does.

Then something actually listened

An AI chatbot. One long conversation, and it started to get me. The more I told it, the better the picks got. For the first time the odds flipped.

Then it started breaking. The thread got endless. The app got slow. It forgot things I'd told it weeks before.

I had finally found something that worked, and it couldn't hold on to what it knew.

So I built something that could.

The first recommendation

The day Ipsomap worked, I asked the chatbot to dump everything we'd ever discussed, poured it into the app, and hit "recommend".

It gave me Killing Zoe. 1993. I'd never heard of it. What I wrote afterwards:

“A real gem. Didn't think I could find something like this from 1993.”

Then Children of Men. Then The Gentlemen. Then Greed.

They weren't made for me. They were found for me, which is better.

The part nobody puts on the landing page

It works because I do the typing.

After every film, a few lines. Not a review. Just the truth:

“focused on emotions and family, nothing to do with Sci Fi”

“credible science and no overrated bullshit”

“I don't think human experience is that important or beautiful.”

And when a pick is wrong, I say why. "Too childish." "Prefer English." "Sounds romantic from its trailer." Thirty-seven rejections so far, every one with a reason.

It still gets it wrong sometimes. But every "no, because…" makes the next "yes" better.

Every second of typing comes back as a good evening.

It told me who I was

I thought I knew my taste. I did, sort of. I just never had the words.

After a hundred-odd notes, Ipsomap wrote this about me:

I read it twice. That's exactly why Margin Call and Zodiac stay with me and happy endings don't. Nobody told me that. I wrote it down one film at a time, and it came back as a sentence.

Sometimes it goes deeper than taste. About DTF St. Louis I wrote:

“Helped me mentally, the depression, the sadness, the imperfection of humans as social beings.”

I would never post that. Nobody else reads it, and that's the point.

There are no followers here, no likes, no public reviews. It's not another place to perform your taste for an audience. It's a place to be honest about it, because nobody's watching.

Nothing gets lost anymore

A friend drops a title at dinner. A film YouTuber I follow trashes something, and somehow now I have to see it. For decades those went into my phone's notes app and died there.

Now they go into my watchlist, with where they came from and a line on why I wanted it. Months later I still know what caught me: "Like that actress from White Lotus and Sex Education."

When I finally get a free evening, I don't scroll. I open the list and they're already waiting, each one with its reason.

Nothing good comes easily

I won't pretend this is effortless. The first months were work: rating carefully, writing notes, explaining every "no".

But that's the deal with anything worth having. You put in the work first. Then it pays you back: films you'd never have found, and words for something you always felt.

And it's yours. Not a trend, not a feed, not what a hundred million other people clicked. A record of what's true for one person.

Funny thing: for years, the film I fell back on when nothing worked was The Matrix. It took me a while to notice I was rewatching a film about exactly this: a system that feeds everyone the same thing, and calls it their choice.

Remember the kid with the spoon?

“There is no spoon.”

There is no good film, either. There's only what's good for you.

Free your mind.

You don't start from zero

You don't have to start where I started. You've probably done more of the work than you think.

A Letterboxd or IMDb export. Your Netflix history. A years-old list in your phone's notes. A screenshot of your streaming queue. A long conversation with a chatbot, like mine.

Bring all of it. Ipsomap reads it, finds every title, keeps your ratings and your notes word for word, and lets you check everything before it's saved. One import, and your taste already has a shape.

Then you start writing.

Whatever your taste is, this is your happy ending. It starts with a few lines of typing, and it gets better every time you do.

— The founder