Hacker Newsnew | past | comments | ask | show | jobs | submit | pousada's commentslogin

I’ve been meaning to build a personal selfhostable version of this. I’ll vibecode it as simple webapp someday and OSS it, should be easy enough.

I see the merit in this project as well but no way in hell i share this data with a third party


> I see the merit in this project as well but no way in hell i share this data with a third party

If get a chance to read the website / privacy policy you’ll find that I actually don’t have access to any data, you don’t need to sign up to begin using the app. The sales pitch is entirely privacy first :) so all the data stays on your device.


You can’t handle the truth

Not on iOS, there Firefox is actually Safari under the hood and you can’t use extensions… Haven’t found a good solution yet (other than avoiding websites with ads)

Wipr2; paid for Safari but it works on all Apple devices with that one payment.

I use reader mode 90% of the time, I’m really not interested in fancy layout or formatting for a website. I just want the text readable and looking exactly the same way for every website. Designers probably hate users like me.

It’s funny, my experience is the exact opposite.

Having grown up in Germany i was surprised how many people in the US would strike up a conversation randomly in the street with me - I thought it’s a normal thing and never really experienced it in Germany - there I would always be suspicious that people try to scam me or get money or something.

I do agree with GP that in Latin America it’s super common and normal to chat with everyone.

But there are many levels to this - it’s for example less common in Nordic countries at least in my experience but you can speak to people in every place on earth, it’s something universal.


I live in a Nordic country. The rule here is to avoid social contact at any cost. People waiting for a bus must stand at least 2 meters away from each other. Otherwise you might be close enough for someone less equipped in the local etiquette to mistake you for someone that may engage in small talk. When sitting , never sit directly next to someone else. Stand if that’s not possible and try to observe the 2 meter rule. Neighbors learn each other time routines to make sure they never have the uncomfortable experience of having to cross paths and acknowledge the others presence.

Regions in the US dominated by Nordic immigrants have similar patterns!


Sounds more like Finland than Denmark. Danish people seem to quite fond of conversations with strangers. I cycled through entirety of Denmark one summer and we’ve had so many beautiful engagements with locals, in villages, towns and cities alike.

I'm on my fifth decade living in the nordics. Strangers have conversations at bus stops and neighbours socialize by the garbage bins and across the hedges all the time.

Not as much as in some other places in the world, but it is not at all rare.


Maybe in other regions, I am in Stockholm and would be shocked by an approaching neighbour trying to talk to me, unless it was to complain about my trees or something.

On the bus/train, even when it breaks down and we have to wait nearly an hour for it to start moving again, I've seen carriages full of people not speaking a word to each other (except people now think it's ok to speak on their phones, to my great annoyance).


As a Pole, Germany was the first place I started enjoying random convos on the street. It definitely depends where as Germans aren’t homogeneous. Hamburg and Schleswig-Holstein are brilliant in that regard. Heck, in Hamburg a dude said my t-shirt is sick and I still remember that vividly. It’s been one of my first solo trips out of my country and it boosted my morale for the rest of that holiday.

Department of War is just such a fucking joke title - when has the US stooped so low, I used to believe in you guys as the force of good on this planet smh

Well then I don't know where you've been for the last ~10~ ~20~ 70 years

When? Its entire history from the foundation of the Republic to 1947. The name was changed after WWII; now a faction wants to change it back. The difference in name never changed the behavior, in either direction.

I'm 33 years old, would you mind telling me which year you thought this was, force of good stuff? might be before my time

genuinely curious, I got nothing


it was before your time.

In WWII, we saved the world from what is now seen as some really evil stuff. Not alone of course, Europe and Russia made huge sacrifices and that's where much of the war was fought. But US arms and blood were the decisive factor, Germany was winning, Japan was winning.

After WWII, the US decided to rebuild the world. We turned our enemies (Germany, Japan) into our close allies.

And the people who did it were really and seriously morally committed to doing what they thought was right. It was about building a country, working together. Not the insane politics of today.

Look, it wasn't all rose-tinted glasses. Bad stuff happened, and McCarthy was worse that what we currently have. And the civil rights movement and all of that. And the stupid wars, Korea, Vietnam, all the smaller police actions. Bad shit was done.

But on balance, the US was seen as the force of good, and the guaranteeor of world peace and the prosperity that allows.


The USA were pretty clearly on the "better side" of conflicts in 1941-1945, during the Cold War (at least as far as Europe and the Marshall plan was concerned). In Koweït and central Europe during the 90s. You may even argue for Afghanistan post 9-11 (although the state building was botched.) in the 2000s. ISIS is a footnote in history because of US intervention (from Trump first term, of all things.) And Ukraine would not be against getting the support it had in 2022 back under Trump.

Does not mean that very bad things were not happening at the same time.

But it's definitely easier to find some "supportable" interventions from the US than, say, Russia or China.


You can’t really because these powerful models are censored. You can create lewd pictures with open models but they aren’t nearly as good or easy to use.

I’ve seen some very high quality NSFW AI video in the last few months. Those models are not far behind and the search and training space for porn is smaller than being able to generate anything

> I’ve seen some very high quality NSFW AI video in the last few months. Those models are not far behind and the search and training space for porn is smaller than being able to generate anything.

Agreed. In my opinion, the primary limitation of the porn models is actually poor labeling of the training set. The company that manages to produce a well-labeled, porn-tuned AI image model is going to absolutely clean up.

The extractive dark patterns that will emerge from a parasocial chat "AI relationship" that can generate porn images relevant to the chat on the fly will be staggering. Once that proceeds to being able to generate relevant video, all holy hell is going to break loose.


> The company that manages to produce a well-labeled, porn-tuned AI image model is going to absolutely clean up.

For anime/non-photographic content that essentially exists (Pony, then Illustrious, then probably some new-fangled thing by now that I don't even know about), thanks to the meticulously tagged booru image corpus. However, as strong as these models are on matters of anatomy and kinks, they're limited in other ways due to the hugely biased dataset and dependence on tag soup prompts rather than natural language (many find the latter a plus, not a minus, though).

I haven't heard of any proprietary/cloud-based NSFW model that would be massively better than what's available for free. There are many NSFW-friendly services, but by and large they're just frontends to models trained by other people.


Because models can be used to alter existing images, you can use open and commercial models together in content creation workflows (and also the available findings of open models, and the ability to further tube them very specific used, are quite powerful on their own), so the censorship on the commercial models has a lot less effect on what motivated people can produce than you might think.

I still think, even with that, that like most predictions of AI taking over any content industries, the short-term predictions are overblown.


Doesn't Grok allow users to create lewd content or did they roll that back?

Also, I suspect that we'll soon see the same pattern of open weights models following several months behind frontier in every modality not just text.

It's just too easy for other labs to produce synthetic training data from the frontier models and then mimic their behavior. They'll never be as good, but they will certainly be good enough.


Just a matter of time and open models will get there. Not once have we seen a moat across the model spectrums.

> Curate the content in your instance

How do i do that without getting PTSD as well? Or is there some magic method that works without me looking at CSAM and gore constantly?


Whitelist instead of blacklist seems like it would work.


How do you know what you can whitelist without looking at it?


Deny by default, allowlist per account. That's what lemmy.ca is doing, you have to apply for an account.

This is speculation; they may look at ips and other fingerprint data to determine if they accept your acc application.


Only allow trusted people to upload your content.


But how do you know who you can trust?


A cursory look through someone's post history.

If a 6-day old account making highly voted posts and no comments? Bot, part of a botnet (check who upvoted, and purge as neccesary).

6 month old account, combination of high and low effort comments? Does not emit hatred with every fibre of their being? Appears to understand debate? Rational human, trust.

You do it on case-by-case basis and slowly increase your trust network


> A cursory look through someone's post history

Thus reopening yourself to the trauma of viewing CSAM.


Whitelist what?


We’re probably a year from self hostable video LLM models that can identify sexual content etc. with high sensitivity (but probably poor specificity)


Yea I can’t see how that could backfire in any way


Quantum computing?


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: