Uploaded October 2026 | Updated November 2026, 3 days ago
Try Hark Pro: hark.com/?utm_source=influencer&utm_medium=social&utm_campaign=slidebean
Hark Pro is an operating system for getting things done. It remembers the context of your life and uses a computer in the cloud to do things for you across the web. Anything you need to do, Hark can handle it. Available on web, iOS, and Android.
--
Check out our own movie recommendation tool to help you find your next great watch.
Try it for free ๐ movies.slidebean.com
--
We pay for Netflix, HBO, and a bunch of other streaming services. They know what we watch, what we like, and even when we give up halfway through a movie.
So why are their recommendations still so bad?
Caya got tired of spending more time scrolling than actually watching movies. So he compared Netflix recommendations across our team, dug into how streaming algorithms work, and discovered that what shows up on your homepage isn't necessarily what you're most likely to enjoy.
Even the thumbnails are personalized to get you to click.
But if Netflix can't reliably recommend a good movie, who can?
In this episode, we break down how IMDb, Letterboxd, and Rotten Tomatoes actually rate movies, why audience scores can be misleading, and whether critics are any better at predicting what we'll enjoy.
Then Caya tries building his own movie recommendation system, experimenting with audience ratings, movie characteristics, and critics whose taste matches his.
Turns out, recommending a great movie is a lot harder than giving it five stars.
And maybe the problem is that taste is just too personal for any algorithm to get completely right.
--
๐ Grab our new Slidebean tees!
๐ shop.slidebean.com
--
Recommended video:
youtu.be/t_GK06_3UY8?si=m3SMtdYFwl0HRPdU
--
Follow us
Slidebean
LinkedIn:
linkedin.com/company/slidebean
Instagram:
instagram.com/slidebean
Caya
Letterboxd:
letterboxd.com/cayahere
Instagram:
instagram.com/caya.here
X (Twitter):
twitter.com/cayahere
LinkedIn:
linkedin.com/in/caya
--
๐ Sources & References
Netflix โ How Its Recommendation System Works
help.netflix.com/en/node/100639
Netflix Technology Blog โ Artwork Personalization
netflixtechblog.com/artwork-personalization-c589f074ad76
Netflix Technology Blog โ GenRec and AI Recommendations
netflixtechblog.com/genrec-towards-llm-native-recommendation-at-netflix-f20be6f643e3
Rotten Tomatoes โ Audience Score Update
editorial.rottentomatoes.com/article/audience-score-update
IMDb
imdb.com
Letterboxd
letterboxd.com
Rotten Tomatoes
rottentomatoes.com
The Movie Database (TMDB)
themoviedb.org
--
#Netflix #Streaming #MovieRecommendations #Letterboxd #IMDb #RottenTomatoes #Movies #StreamingServices #Algorithms #AI #MovieRatings #FilmCriticism #Slidebean #HarkPartner
Try Hark Pro: hark.com/?utm_source=influencer&utm_medium=social&utm_campaign=slidebean
Hark Pro is an operating system for getting things done. It remembers the context of your life and uses a computer in the cloud to do things for you across the web. Anything you need to do, Hark can handle it. Available on web, iOS, and Android.
--
Check out our own movie recommendation tool to help you find your next great watch.
Try it for free ๐ movies.slidebean.com
--
We pay for Netflix, HBO, and a bunch of other streaming services. They know what we watch, what we like, and even when we give up halfway through a movie.
So why are their recommendations still so bad?
Caya got tired of spending more time scrolling than actually watching movies. So he compared Netflix recommendations across our team, dug into how streaming algorithms work, and discovered that what shows up on your homepage isn't necessarily what you're most likely to enjoy.
Even the thumbnails are personalized to get you to click.
But if Netflix can't reliably recommend a good movie, who can?
In this episode, we break down how IMDb, Letterboxd, and Rotten Tomatoes actually rate movies, why audience scores can be misleading, and whether critics are any better at predicting what we'll enjoy.
Then Caya tries building his own movie recommendation system, experimenting with audience ratings, movie characteristics, and critics whose taste matches his.
Turns out, recommending a great movie is a lot harder than giving it five stars.
And maybe the problem is that taste is just too personal for any algorithm to get completely right.
--
๐ Grab our new Slidebean tees!
๐ shop.slidebean.com
--
Recommended video:
youtu.be/t_GK06_3UY8?si=m3SMtdYFwl0HRPdU
--
Follow us
Slidebean
LinkedIn:
linkedin.com/company/slidebean
Instagram:
instagram.com/slidebean
Caya
Letterboxd:
letterboxd.com/cayahere
Instagram:
instagram.com/caya.here
X (Twitter):
twitter.com/cayahere
LinkedIn:
linkedin.com/in/caya
--
๐ Sources & References
Netflix โ How Its Recommendation System Works
help.netflix.com/en/node/100639
Netflix Technology Blog โ Artwork Personalization
netflixtechblog.com/artwork-personalization-c589f074ad76
Netflix Technology Blog โ GenRec and AI Recommendations
netflixtechblog.com/genrec-towards-llm-native-recommendation-at-netflix-f20be6f643e3
Rotten Tomatoes โ Audience Score Update
editorial.rottentomatoes.com/article/audience-score-update
IMDb
imdb.com
Letterboxd
letterboxd.com
Rotten Tomatoes
rottentomatoes.com
The Movie Database (TMDB)
themoviedb.org
--
#Netflix #Streaming #MovieRecommendations #Letterboxd #IMDb #RottenTomatoes #Movies #StreamingServices #Algorithms #AI #MovieRatings #FilmCriticism #Slidebean #HarkPartner










