honestly, depending on how it affects media in the future, it could really have a similar effect on us, even if it doesn’t directly poison people. Poorly presented information/misinformation has, and will certainly continue to result in so many preventable tragedies.

search results have absolutely died ever since last year honestly, and each month it feels even worse somehow.

AI seems to use some weird black-box approach to find sites nowadays? That might be the reason why my stuff is hit with so much random garbage nowadays.

I can understand this viewpoint (and it’s similar to the first point), and I do agree that there is a push to defer things to AI. But from my limited experience, it’s like a crutch for other issues that force it to be used by very talented people who otherwise want to get things done properly to save headaches later.

My experience was in the kind of place you would really not want AI to screw you over in, but there was also a very tight timeline to get things done, so I think they ended up just trusting AI to a pretty frightening extent with (iirc) some buildroot stuff that got some people a bit vocal. I think even I used an AI overview once because I just really wanted to be done with a demo.

It was over a year ago though so I don’t really remember what ultimately came of the AI use, but there are 100% situations where this becomes the only option, although I personally will avoid AI unless I’m realistically out of any other option not to. And for that reason I’ve fallen more out of favour with CS since I enjoy the journey and all the little judgements you make more than the outcome.

Computer science, and quite scarily, a bit of medicine?

Pharmacists are gonna have to review even more slop now with this scope creep smh /j

AI/LLMs replace cognitive work while the outcome is not guaranteed.

Calculators… are devices designed to produce deterministic results that can be proven (and yes, with a seed AI is technically deterministic, but that doesn’t make it easily provable) as accurate very easily. If you removed calculators from a task, either the result would be completely the same, or it would be worse due to error introduced by something unpredictable.

past like really, really early education, I don’t really see a point to go without a calculator since it uses up so much time for little gain.

I mean you probably could do a decent range of web apps in a day as a minimum viable product, if you already have a solid plan ready. But that’s kind of the thing with AI.

AI… really seems to just make derivative stuff. I tried asking it to replicate one of my terminal text editors (via a description of it), and it ended up making… essentially a 1:1 rip off of different editor but with more comments. I have a feeling that the massive amount of training data is what makes it seem like it can write lots of applicable code, while it’s still suffering from quite literally being just a larger version of this half-decade old Google Translate update under the hood.

well if the bots do implement checks for this software, that also wastes their compute, but at a disproportionately large level just to make sure a few images (compared to actual billions of other images) aren’t obfuscated.

That would increase complexity, but I kind of like the ability to gauge what the image was probably supposed to be. It seems like AI can’t identify objects as easily even when the changes aren’t really that drastic.

Currently I am experimenting with using a ‘phrase’ to encrypt the image in chunks as well, which has shown some really promising results!

Anubis can do an okay job… (at least so the whole site isn’t being actively DDos’ed) at the cost of compromising user experience a lot by sending harder challenges more often. It kinda sucks that the average user might not have that fast of a CPU, while motivated scrapers might spend tonnes on compute just out of spite.

(i am a bit biased on this though)

you can use it to (sometimes) transplant your personal config and documents onto another OS, but it comes with the tradeoff of losing space in root that might be used for massive libraries (compute libraries suck in this regard).

And sometimes your /home will depend on things that your rootfs doesn’t have, which has caused problems for me before. But if you planned for these things beforehand, and your installed software doesn’t do something really dumb, it can work out pretty well, although my personal experience with this is limited to effectively embedded Linux where this is usually a non-issue.

Even ext4 still has safety measures (BLOAT!11!!1) because of journaling.

This is why i compile my kernels on ext2 with without swap for maximum utilization /j

I wonder if using a hash-based system could work slightly better (on platforms where you can’t edit image links) in these cases where a platform decides to stop serving to some users or go defunct. So that way, you can at least have a slight chance of ‘fixing’ an image after the fact.

It wouldn’t work with most image share sites, but at least you could rehost it on a mirror yourself.

if we are being kinda serious, we would probably just go with either the systemd user metadata solution, or… (on a scale of insane to less insane):

• store it in a passwd comment (which means the finger command will now remotely leak age, but is backwards-compatible)

• have an SELinux (or similar) setup so that new users have more restricted permissions by default. While I would absolutely do this myself whenever I have kids, this is just really unnecessary for most everyday users. At the very least this obfuscates the reason why a user might have less permissions.

• make a generic, encrypted-by-owner file in /etc/, but this still has privacy issues if we need every account to state their age.

Obviously the best solution is just educating parents about managing parental controls, but we can’t have nice things ig.

for once this is just me reverting back to caveman brain where i just do > char + 64 and char + 96 tests so I can avoid stdlib on C… (which is really dumb)