Oh yeah sure it’s not the ethics thats actually the problem. I think everybody agrees on the need for a strict ethical framework.

But most of the research institutions that I have been involved with have cared very little about the actual ethical constraints of research (such as data privacy or survey questions that could be triggering) But every single time they will pull you up on the font being too aggressive, whatever that means.

I can’t speak for regions other than my own however.

The bureaucracy of a typical ethics review is insane and it neither helps design ethical experiments or set boundaries, it’s just paperwork concerned with font type. there’s more truth to this than we’d like, which is not ok.

Well that’s great, airvpn has worked well for me in my torrent docker container and I recommend it for that purpose.

For privacy I recommend arkenfox

I believe airvpn has port forward.

Well it’s probably both.

Insights from the data probably were not being actioned, but I would strongly suspect that the data they are collecting simply doesn’t have a lot of predictive capacity.

However, I don’t work with that specific data. I work with related data.

The Western world is far too concerned with test scores when they are just complete and utter bs.

I would say a test score is accurate plus or minus 40 % in terms of a student’s understanding of a subject. They’re just arbitrary.

I work with school data, I’ve worked with University, high school and primary school data. it is indeed all bullshit I’m large part because test scores are just noise and behaviours metrics are subjective and non standard.

I’ve never been able to develop a model with any predictive capacity whatsoever at all. Moreover, visualisations only ever show correlation and often do more harm than good as staff assume their actions are causing improvement when typically advantaged students simply take advantage of more activities etc.

The post above is certainly more insightful than Elon musk’s opinion and this is coming from somebody who works with this type of data.

Again, I wouldn’t suggest pulling it all apart. I would look deeply into the problem but this is really not the worst thing they’ve done.

The whole notion of LSP has been nice.

I use it a lot when I’m writing my notes (ie joplin/obsidian), I’ll use flux or stable diffusion for a few iterations until I can create an image that Is consistent with what I’m writing.

It can be really convenient to be able to recognize an image as you’re browsing through notes that are otherwise just filled with code or maybe a recount of the day.

I’m sure most consumers consider excessive use of generative AI to be in bad form. It certainly doesn’t exude professionalism.

Oh, okay, I understand what you’re saying now.

Yeah, I don’t trust any of the VPN providers. There’s just no evidence that they’re trustworthy. I reach for Tor (or i2p sometimes).

I typically run all the torrenting stuff in a container, I’ve never actually used that VPN to browse. I just spin the container up and down when I want my bandwidth back.

I’ve had a good experience with AirVPN. I mean, I only use it for torrenting, but… Is there a good reason not to go with them for torrents?

Keepass with rsync / unison or a local git server works pretty well too.

Airvpn has port forward i believe.

I don’t actually dislike ai imagery, I think it can produce interesting imagery. However, I must concede that is an excessive use of boilerplate bog-standard AI imagery.