Yes, anyone who has been looking for Spotify alternatives can readily tell how poor the music recommendation algorithms of competitors. I feel like the only comparative one is weirdly, youtube, since they simply have more data points and context to attach recommendations to.

Seems like you don't have to rule it out completely? "You might also like this show" is rather different than YouTube's "OMFG SHORTS WATCH THEM NOW" experience. Even if the line is fuzzy, I think many can agree some things are closer than others, and that's enough for some laws (there are plenty of other not-perfectly-testable laws out there as counterexamples).

> I want those things. I make use of them.

Can you opt out of those features? If not, how do you know you want those things or if you are just being manipulated into using them?

  > when I have exactly 2 hours of TV watching time in a week and don't want to waste it finding a new show.

The television is a tool for watching shows. The show is ostensibly the draw, and the television is the vehicle for watching it, no? So why do some people (lots of people actually) feel compelled to sit in front of the television to watch something? If there is no show that one wants to watch, why the compulsion to sit in front of the television?

I've spent countless nights in front of Netscape Navigator surfing from site to site. I'm fully aware how compelling the device-content system can be. But it's always been the content side, not the device side, that was compelling.

There's a big difference between a completely optional "recommendations" area and a loop. If the recommended stuff happens automatically without you having to look for it, then it's a loop.

Ok then at the very least allow for a toggle to turn those algorithmic “recommendations” on and off

Cigarette packages have disgusting photos of carcenogenic lung autopsies to discourage you from smoking them, so Facebook and TikTok etc should have similar disgusting photos and vidoes of what happens to you if you use their product, that you have to watch every day before using it.

Sleep-deprived teenager autopsy: a teenager with enormous purple eye bags, phone surgically fused to their hand, lying face-down on a school desk at 3 AM.

Body-image filter ward: rows of teenage faces grotesquely stretched, pinched, inflated, beautified and looksmaxed until everyone looks like the same embalmed Instagram mannequin.

Doomscroll thumb: a gigantic swollen thumb with the fingerprint worn completely smooth from infinite scrolling. The tendon is labeled "ENGAGEMENT."

Facebook scam victim: an elderly person's wallet dissected on an autopsy table -- completely hollow, with a Facebook ad where the money used to be.

Algorithmic stomach contents: an autopsy tray containing "suggested for you": rage bait, diet videos, conspiracy sludge, influencer ads, and one photo from somebody you actually know.

Predator warning: a cute Instagram DM screen shown as an anglerfish -- friendly profile photo as the glowing lure, enormous teeth hiding behind it.

Facebook Hospital: a free to play educational game where you have to spend virtual currency for cures to the ailments Facebook gave you, like Bloaty Head from reading all the AI generated engagement porn.

https://www.youtube.com/watch?v=Le_znuXcP2M

> > “Do you use personal / behavioral data / usage to shape the user experience and recommendations, creating a reinforcing loop?”

> This would include everything from my Spotify recommending new music that I like to my Netflix showing me similar shows.

Spotify recommending music and Netflix recommending shows that are similar to what? To previously recommended music/shows. This is exactly the problem.

> I'm a responsible adult. I want those things.

This is a clear indicator that these things are addictive and are in fact gambling.

Law is about intent. It is not code, and it’s a mistake to think of law like code where you need to unit test and cover all edge cases.

In law, it is sufficient to show an intent to take actions that foreseeably caused harm. Specific laws, of course but you don’t need to cover hypotheticals that capture all abstract cases in order to to show a particular actor caused harm.

You’ve not already noticed that the end of every consumer-facing LLM chat always includes an invitation for just one more turn to keep the loop going?

> Okay but suppose you have an AI chatbot which takes in context from previous chats on what your interests are. So people just keep talking with the chatbot more and more as it gains more context. This would meet the criteria. It's not necessarily addictive though because you can stop talking with the chatbot at any time.

You can stop scrolling instagram "at any time", too.

Chatbots are actually a pretty poor but very relevant choice of example: they are already tuned with RLHF to be engaging and addictive to talk to. That's a huge reason why LLM psychosis is a thing, and it's a big reason why some people like using chatbots so much. They're designed to be addictive, just like facebook/instagram/tiktok are.

"You can stop smoking at any time"

"You can stop gambling at any time"

As much as I may agree with you, you need a different argument.

I'm not sure about that example. I'm seeing non tech people getting very addicted to their customized chat bots. Companies are very primed to eventually nudge them to enshittify the content of the responses and it could get bad.

> It's not necessarily addictive though because you can stop talking with the chatbot at any time.

You can also stop taking drugs at any time. Or stop drinking. Or stop smoking. Or stop doomscrolling.

Suppose you have a video service which takes context from previous videos you watched to show you more videos you're interested in. Doesn't sound very addictive, does it? You described it in an innocuous way, but I can absolutely see that being a lead-in to a parasocial "relationship" or even AI psychosis.

> you can stop talking with the chatbot at any time.

doubt. Lonely people will get addicted.

In fully supportive of prosecuting wrongdoing in the food industry.

But it seems strange to use belief in widespread wrongdoing in the food industry to oppose punishing wrongdoing in the tech industry.

Is it just nihilism? All industries are corrupt, therefore prosecute none of them?

why do you think that implies leaving out the modern food industry, rather than suggesting that they, too, should be held accountable for flooding us with hyperpalatable foods to addict us? After all, they were literally engineered by the same people as Big Tobacco [1]

[1]: https://www.washingtonpost.com/wellness/2023/09/19/addiction...

Unlike big tech, which would never do that.

A hammer doesn’t have a sensor to know how you are doing while you're hammering away, and it doesn't make the strike smoother to keep you hammering longer if you start seeming tired. Almost no products ever monitor you

Most products don't spy on you?