How to not seem as a sell out, or to put it more nicely: How to monetize academic expertise without losing credibility?

Oli mennyt itseltä ohi mutta hienoa, että siitä levitetään tietoa mediassa 😊 Auringon kukat voisi toimia viittauksena myös Ukrainan tuelle.

Oliko tämän argumentin tarkoitus puoltaa ydinvoiman ympäristöystävällisyyttä?

Ihana Katariina! 😍 Tuo mekko on tietenkin ilmeinen viittaus Hidden Disability -nauhaan, jolla viestitään näkymättömästä vammasta. Katariinalla on ko. nauha usein kaulallansa 😊

I admit that my time on elementary school was horrible so most of my memories from there are unpleasant, so it's possible that school association is affecting my attitude towards the book. Perhaps it would have been more pleasent reading experiment if my mom would have read it to me :)

I was made to read those books in elementary school and I thought they were deadly boring... The hidden satire was lost on kids and I don't remember anyone was exited about them. Even today the title of the book brings back unpleasant memories. Perhaps the book was meant for adults after all.

Miun YouTube Musicin Recap näyttää koostuvan suurimmaksi osaksi virolaisesta popista 😄

I wonder what is the price that they get to play with that data...?

hashlytics.io/genesis-miss...

Vähemmän häiritsee kova pakkanen kuin kova helle, mut mieluummin toki +24 kuin -24. Talvella on kuitenkin kiva, että olisi pikku pakkanen, jotta pääsisi nauttii talviharrastuksista :)

💔🕯️🥀

Hauska peli, kiitos vinkistä! 😊

Neloset 14.11.2025
6 yritystä (pelaajien keskiarvo 6)
2/3 vihjettä käytetty
🟧 🟧 🟧 🟧
🟦 🟦 🟦 🟦
🟩 🟩 🟩 🟩
🟪 🟪 🟪 🟪

www.is.fi/kotimaa/art-...

If I manage to demonstrate this phenomenon even in a simple simulated setup, that's already enough reason to be alarmed and to look for better solutions for Reinforcement Learning based Recommender Systems.

It sounds like you're criticizing the simulation research method in general, are you?

2/2

First of all, the focus here isn't on human behavior but on RS behavior. I don't need perfect human modelling (would prefer tho) to show that some RSs (RL-based) tend to manipulate users. They have a clear incentive to do so, and I want to show that they can unnaturally shift user preferences. 1/2

In my thesis I'm demonstrating how certain types of recommender systems have a tendency to change user preferences, to manipulate them. It would be highly unethical to use real humans to show that indeed their preferences can be unnaturally shifted. Simulation-based approach is the next best thing.

Would be interesting to know if LLM would model human decision making better than the Markov Decision Problem -method (or some version of it). Guess I have to find out! Thanks for sharing that literal review 😊

That's very interesting! What led you to this topic? I'm writing my master thesis about manipulative recommender systems and for it I need to make some simulation where I demonstrate the RecSys-human interaction. Haven't yet figured out how to simulate a human user but this could be one way to do it

I'm thinking about switching to Nextcloud, but after 20 years of using Google services it's easier said than done! It's going to be a long process to move all my data...