

Everything AI does is with stolen data. What makes this so special?


Everything AI does is with stolen data. What makes this so special?


Well, the bots do. Like 90% of viewbotting, click farming, follower buying etc is sold by people running phone farms.


They are separate clauses in that sentence. Please learn to read.


ok well, the people who use github and write programs are probably a lot more likely to have experience with claude than you are so they probably didn’t make the same assumptions you did!


Because most people have at least tried these products and it’s not like the agent is being hidden away. It makes commits to your repo using its name as the author (for now).
Do you assume anyone who knows something you don’t has no life? Strange way to live.


cpython is the reference implementation of the python interpreter. The person who took this screenshot has the Claude user on GitHub blocked so that whenever it contributed to a git repo you see this warning. The Claude user is an AI agent. AI code is garbage.


There’s a giant, glowing, animated LED billboard along a main road near my house that had a PSA about distracted driving on it the other day. It made me angry.


deleted by creator


I’m not familiar with the exact techniques used but I am aware that there are ways of fingerprinting images that would survive such transformations.


a lot of the big names in image generation fingerprint the images they output to make them easy to filter out of training data. theres a possibility they could detect this fairly easily.


That would work if you were rich and friends with government officials. I don’t like your chances otherwise.


That’s fair. I just assume most people who reflexively insult JS are like that usually because they’ve written something in it, rather than having dealt with the many vulnerabilities and annoyances browsers have made possible with it


Browser permissiveness didn’t create implicit type coercion.


Not as hard, I agree but reimplementing just CSS sounds like a very special level of hell.
Then run it in a container under a better distribution if you desperately need to put neofetch on your HTPC. Or run the other distro in a container under libreelec since I’m pretty sure it supports them.
It tracks anonymous statistics, without my express consent, for the benefit of a third party. I do not care if it exists to replace cookies, because I’m not even convinced that cookies need to exist at all anymore. What utility do they provide to the actual person using the browser that can’t be accomplished through some other more modern API? If the only functionality left to replace is tracking people then maybe just deprecate them and move on.
Telegram had credibility. It was being used by journalists to protect sources.
You can extend trust to individuals but do not apply that to companies or organizations if you care at all about what they’re doing with what you give them. Not everyone has some mythical tech privacy wizard on call to give them perfect advice every time they open an account on an app or website.
Even client side encryption is not infallible. The algorithm you use will eventually be crackable and probably sooner than you think. Nothing lasts forever.
The most foolproof way to ensure something remains private is to not put it on the internet at all.
If you can read and understand the code, sure. Otherwise you’re still just extending trust to someone perhaps less reputable than even the corporations who are dying to sell you out. For example, the back door some mysterious contributor slipped into xz recently.
My recommendation is to live life as if privacy on the internet did not exist, because it doesn’t.
Never trust a third party to keep your shit private. Especially if privacy is their main selling point.
“Don’t be evil” seems like an aspirational sentiment rather than a reflective one. Like an addict writing “don’t do drugs” on their todo list.