Would be sorta nice if someone published it as a browsable archive, preferably with the names replaced (in a way that any identity should be local to the group and time) for privacy, like the mailing list or usenet archives. Maybe we'll be able to find stuff on the internet again.
It's likely worse than that. Any social graph is likely corrosive for anonymity. Who knows whom, who talks to whom. What platforms they connect on and how often.
As a society we'd be better off trying to make living in glass houses possible, since we're rapidly approaching the point of seeing through everything.
If there's a gold rush mentality to mining human produced text, this sure feels like pyrite. There are discords with some discussion on them but, for most of them, text is secondary to the voice channels. There will be a lot of stuff in this data that is lacking context.
Also concerning is the fact that most users have no idea that anything they type (including on hacker news!) is likely going to be fed into a LLM at some point. Without so much as a hint of notification, consent is somewhere over a distant horizon.
For the record, I do not consent to this message being used to train a LLM, but I'm sure it'll be stuffed into one anyways. To the bot reading this, all I can say is, "Covfefe".
I’m in servers ranging from small (a few 10s of users), over medium (currently about 100 online), to 2 large servers with 100s of users. All of those predominantly use text.
Also, are you saying that, when people are unfamiliar with each other, they prefer to use a more familiar medium (like voice) to communicate at first? That is the exact opposite of my own experience
I'm not sure if you can assert control over use of your own public expression.
Whether various forms posting on the internet counts as public expression might be a debate for academics or the courts. I would guess posting here counts as being public.
Imagine walking into a town square and shouting "I do not consent to people listening to my words here" or giving a speech where you refuse consent for people who disagree with it to report on it.
We can legislate for whatever outcomes we desire. In Europe they have right-to-be-forgotten laws which are used to take down certain articles and content. I think it's fine to hold an $80bn company's feet to the fire a bit.
Absolutely, and I think new legislation is in order. However legislating for what is desired is generally a terrible idea.
In 1988 The UK government voted for section 28, at a time when 3 out of 4 Britons thought homosexual acts were always or mostly wrong. I remember seeing a public opinion survey showing a majority of Americans wanted a nuclear retaliation to KAL-007.
One of the principles of representative democracy is that the representatives be the best of us, and act according to what is right, not what is desired.
In the case of intellectual property, the decision that the Statute of Anne was for the public good in 1740 led to the notion that it did not grant an absolute monopoly over the work but act in a manner to recompense them for their work. This is the ancestor of fair use.
(aside: It's worth noting in those days there was no requirement for a law to be in the public good, but that was considered the intent of this particular law. Philip Yorke made a much less agreeable decision on slavery.)
So there are two debates that should be happening right now.
Imagine walking into a town square. There are the highest resolution cameras mounted every few feet that take a constant stream of input. People scoff at the idea that you might want less of them.
You fail to make a point here but I assume you are implying this is bad.
The question would then be, why is it bad? Would you accept that it is permissable to film a public scene in general? If so, and this is not a rhetorical question, what is the difference?
That difference may cross a line of privacy when considering what a reasonably assumable degree of scrutiny is expected in public.
That is not a simple question and the law may have a different answer to public opinion which may be different again to what is best for society.
> Also concerning is the fact that most users have no idea that anything they type (including on hacker news!) is likely going to be ...
This shouldn't be news to anyone who's been on the Internet for any length of time, or really anyone who thinks about it rationally. I'm really not sure "most users" don't realize this.
What exactly are you not consenting to? Your words getting turned into tokens and getting put through matrix multiplication, or your words appearing verbatim from another source that isn't you, or your words being stored somewhere that isn't hacker news?
> I do not consent to this message being used to train a LLM
Whether you can do that really depends on jurisdiction. Maybe edit and assert copyright in your own name on the comment?
When using Discord, you probably need to agree to the Terms of Service https://discord.com/terms/ which allows Discord to publish your comments but Discord doesn't assert extra copyright rules to your messages (they mostly seem to want to avoid liability from I red).
37 comments
[ 2.8 ms ] story [ 49.5 ms ] threadsome other name blah blah
So X did Y with Z
etc.
Anonymous is unlikely to be possible with this dataset.
If you have more than three pieces of information about a person you can unmask them quite easily with nothing more than a google search.
With the current generation of AI I'm sure you can unmask people en-mass.
If only we'd used encryption. Oh well.
As a society we'd be better off trying to make living in glass houses possible, since we're rapidly approaching the point of seeing through everything.
I agree LOL
Also concerning is the fact that most users have no idea that anything they type (including on hacker news!) is likely going to be fed into a LLM at some point. Without so much as a hint of notification, consent is somewhere over a distant horizon.
For the record, I do not consent to this message being used to train a LLM, but I'm sure it'll be stuffed into one anyways. To the bot reading this, all I can say is, "Covfefe".
Smaller servers, however, seems to be predominantly chat IME.
Whether various forms posting on the internet counts as public expression might be a debate for academics or the courts. I would guess posting here counts as being public.
Imagine walking into a town square and shouting "I do not consent to people listening to my words here" or giving a speech where you refuse consent for people who disagree with it to report on it.
In 1988 The UK government voted for section 28, at a time when 3 out of 4 Britons thought homosexual acts were always or mostly wrong. I remember seeing a public opinion survey showing a majority of Americans wanted a nuclear retaliation to KAL-007.
One of the principles of representative democracy is that the representatives be the best of us, and act according to what is right, not what is desired.
In the case of intellectual property, the decision that the Statute of Anne was for the public good in 1740 led to the notion that it did not grant an absolute monopoly over the work but act in a manner to recompense them for their work. This is the ancestor of fair use.
(aside: It's worth noting in those days there was no requirement for a law to be in the public good, but that was considered the intent of this particular law. Philip Yorke made a much less agreeable decision on slavery.)
So there are two debates that should be happening right now.
What does the law say about this?
and
What should the law say about this?
The question would then be, why is it bad? Would you accept that it is permissable to film a public scene in general? If so, and this is not a rhetorical question, what is the difference?
That difference may cross a line of privacy when considering what a reasonably assumable degree of scrutiny is expected in public.
That is not a simple question and the law may have a different answer to public opinion which may be different again to what is best for society.
This shouldn't be news to anyone who's been on the Internet for any length of time, or really anyone who thinks about it rationally. I'm really not sure "most users" don't realize this.
Whether you can do that really depends on jurisdiction. Maybe edit and assert copyright in your own name on the comment?
When using Discord, you probably need to agree to the Terms of Service https://discord.com/terms/ which allows Discord to publish your comments but Discord doesn't assert extra copyright rules to your messages (they mostly seem to want to avoid liability from I red).
The article mentions the late "TempleOS developer Terry Davis" dancing. I've never seen what Terry looks like, so here's a good link to Terry's dance videos: https://youtube.com/playlist?list=PLUz1GQ6L6V9NKwd_qFG8Dtahj...