Ooooh...

I love the art, but I don't envy your nightmares...

Yep, last year I didn't get the vaccine because the £100 it costs here seemed excessive, but then ofc I got COVID and couldn't take extra work because of that. If I got free vaccination, I would have likely paid that back in taxes from that missed income alone.

I've learned my lesson

Do you know what went wrong? Are they lacking a good process, or do you think it's malicious?

Our first stop: Housing Policy.

Why is rent so high? Is the crisis actually being solved anywhere? And what does the crisis have to do with wealth inequality?

Katz takes us to Vienna, Finland, and Paris in search of the policies that work.

www.europeanspodcast.com/all-episodes...

My condolences

Oooh, I love the mood!

Round birb 🖤

You tricked me not only into following you, but also reading and enjoying Book of CoL 😱

I once saw a guy alone on a tandem bike, carrying a bouquet of flowers and to this day I sometimes wonder what was the story behind it

I love them! The labyrinth is intriguing

Now these platform have enough tech and compute to train a single recommender system using all their users data. So, they better guess what's similar to what you already like, but are less likely to "by mistake" recommend you sth new, and will be biased towards the average user taste

But, there were many companies that wanted to recommend to you stuff that will keep you the longest on the platform, and one of them was Netflix, who created the famous Netflix Prize for the best recommender algorithm (that's the only connection to Netflix, yes it was clickbait)

I'm _guessing_ that they chose that partly because a) it was a good heuristics that people would like stuff that "similar people" like and b) because it was easy to optimise computationally, you didn't have to have huge clusters or burn rainforests

For example, ages ago @last.fm let you listen to your "musical neighbourhood", i.e., to music that people with similar music taste listened to. It was almost like getting recommendations from talking to people during a concert break (not that I did that a lot, I'm a nerd).

In my educated guess, one reason why the recommendations by these "recommended systems" seems less interesting, not pushing you to explore new music is that machine learning algorithms got better and compute cheaper.

I like this one! 💚 Could see it turned into a pattern for s fabric or something