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

I'm guessing someone saw me post this link in the muJS post earlier, which I got massively downvoted for. And then again when I said I didn't want to use github because of the AI training.

Your comment breaks multiple site rules... please don't.


This is bad advice to a new FLOSS project that wants to have users. Avoiding GitHub with its user base (meaning issues and discussions), search, project topics (tags), trending repository lists, etc. will make a fledgling project even less likely to gain adoption.

A better thing to suggest is to use multiple forges, including GitHub, and mirror your projects across them. This way you will have exposure and options; you won't be as tied to any one forge.


Hard disagree, multiple forges does not solve the problem of being unable to opt-out of AI training from your code.

If that is your problem with GitHub, then I agree, you should avoid GitHub, though someone can still mirror your repository there. I assume most new FLOSS projects that want to have users don't consider it a dealbreaker.

If your code is in any way public, it will be trained on. That ship has already sailed.

If your problem is with your code appearing in training data, then you cannot release your code anywhere.

That link you provided only points out GitHub has integrated "create pull request with Copilot" that you can't opt out of. Since anyone can create a pull request with any agent, and probably is, that's a pretty dated complaint.

Frankly not very compelling reasons to ditch the most popular forge if you value other people using/contributing to your project at all.


Thanks for the link. I'm aware of the debate around GitHub. For now it's where µJS lives, but noted.

It is not worth taking seriously people who have very niche complaints that don't even use your software, believe me. It's the same as some people here who complain that they disable JS and some web app doesn't work, like yeah, no shit. Catering to these people will just frustrate you and make you more likely to give up creating the product.

The mods do not seem to care about pirated software (or links to tons of it) being posted here, I have seen people doing it for years now.

Animal cooperation proves that game theory is universal, but it does not prove it works in a vacuum for humans.

- Biology gives us the instinct to cooperate and the capacity for empathy.

- Capitalism provides the mechanism to scale that cooperation to millions of strangers.

- Institutions (laws/culture) provide the rules that prevent the "vacuum" from devolving into a state where the strongest exploit the weakest (which is actually what happens in nature when policing fails).

Therefore, in a capitalistic society, cooperation to the detriment of the individual (e.g., paying taxes, following labor safety rules) is not just a biological imperative; it is a social contract enforced by culture to allow the complex system to function. Without the cultural layer, the biological layer alone is insufficient to sustain a modern economy.


I am logged into way too many sites to do that unfortunately. I do use a password manager with a browser plugin to make it easier, but it's still a lot of manual work to re-login to all the sites I use on a normal basis, for both work and home, every time I restart my browser.

Would be nice if there was some other solution, like maybe encrypting the browser profile and then requiring a pin/password/biometric/something to unlock it on each start.


There's a Cookie AutoDelete plugin [1] that cleans up cookies, cache, etc for a site after all of its tabs are closed. You can exclude sites that you want to stay logged in to.

[1] https://github.com/Cookie-AutoDelete/Cookie-AutoDelete


It shouldn't take more than one second to log into a website using the Firefox password manager.

In my case it often can and does.

Many sites I use force email or SMS-based 2FA, sometimes in addition to "security questions" and/or have other multiple steps of authorization (like captchas) required; it's often not just a simple username/password for me.

Now multiply that by 25 different sites. Not happening.


One option for that is to use multiple Firefox profiles. The main general-purpose browsing profile would have a hardened configuration, while dedicated profiles are used for other websites that should remain logged in.

There are but I'm not aware of anything that can reliably fool creepjs.

https://abrahamjuliot.github.io/creepjs/

And yes it often results in endless captcha loops.


Source:


I read the filings.

Nowhere in that entire case does anyone allege that the FBI was regularly being sent entire copies of the hard drive contents of best buy customers.

The FBI merely taught workers how to identify and report CSAM. There is nothing illegal about that.

EFF only sued because their FOIA request for info about their training process was denied, and after the FBI argued why they shouldn't grant the request, EFF agreed and backed down.

Not only did the EFF agree to dismiss the case, their blog post claim of a supposed Fourth Amendment violation was never even argued in any of their filings at all.

In my opinion, to construe a simple disagreement/misunderstanding over a FOIA request denial (which was proven as legal and justified) as "If you took your laptop to Best Buy for repairs, the FBI got a copy of your hard drive contents"... is patently and demonstrably false, and does not make any sense whatsoever.


So you think in this case the EFF was wrong? It seems that way, but I'm not sure I fully understand what you meant. Why wouldn't the training process be public?

Another thing is that while perhaps entire copies of customers' hard drives weren't sent to the FBI, the Best Buy repair staff dug through the contents of people's hard drives. If I have a software issue with my OS (or whatever the repairs were about), I wouldn't expect the repair staff to look at my photos. Obviously, is CP was set as the wallpaper or something, you can't miss it, but why is it OK to look into random folders looking for suspicious files?


You’re making the same mistake the EFF’s post wants you to make.

Employees were trained on how to identify and respond to CSAM. The training material was not released based on the FOIA request.

That doesn’t imply that the employees were poking around above and beyond where they had to look to do their job, and it doesn’t imply that full copies of your hard drive are being copied to the FBI.


Ok. I didn't make the claim and I'm not arguing this with you. You asked for sources and I assumed good faith. I was mistaken.

nope, and I sometimes walk with a pebble in one or more shoes /s

If you walk without rhythm,

Yeah my first thought was "of course an LLM can do that, we didn't need a paper to tell us". I would be more impressed if it could do it without that information, such as by analyzing writing styles and other cues that aren't direct PII.

It’s the same thing as theft and locks. Any motivated attacker will overcome any rudimentary obstacle. We still use locks because most opportunistic attackers are the most prevalent.

Even the paper on improved phishing showed that LLMs reduce the cost to run phishing attacks, which made previously unprofitable targets (lower income groups), profitable.

The most common deterrent is inconvenience, not impossibility.


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

Search: