Aspie96

Article:

Digg isn't going away.

Post:

Digg is gone again.

Japan changed its flag for Christmas…

utterly unnecessary

Is your alternative another versioning software or just a massive pile of folders?

I wonder if this includes pure donations. As far as I know (INAL, yada yada), you don't need to pay taxes on pure donations in Italy.

Would it be illegal for me to place a Monero address on my blog to receive small (anonymous) tips?

What's a "free speech person"? Is it one that support effective access to the actual exercise of fundamental human rights?

Yes, the original tweet with a magnet link seems like a good source.

I don't think it's hyper-individualistic. In fact, it's the opposite of that. It groups people by random characteristics and then it describes "identity" as the collection of these characteristics, rather than something unique to each of us.

It's anti-individualistic, if anything.

Indeed there is not, but there is such single thing as "United States" (even though thee are multiple states) and certain forms of "activism" are specific to the US, or at most the anglosphere.

What Europe has is that it's Western, but it's not the US.

Over here in Europe, people have a much more relaxed approach to AI safety.

And to most things, to be honest.

Politics in Europe (I'm Italian) is nuts. But it doesn't even remotely approach the nut-level of the USA.

Outside the English-speaking world, identity politics didn't really take hold.

To some extent it does, but only as indirect influence from the anglosphere. In fact, most of the linguistic games Americans like to play to pander to one or another group don't translate well, if at all.

An European leftist is much closer to an American leftist of 2010 than to one of 2023.

European leftists might be much at the left of Americans when it comes to economics. An Italian right-winger is probably at the left of an American left-winger. Socially, the focus tend to be on actual issues and actual discrimination. Trying to continuously change the way language is used gets nothing but the ridicule it deserves. Meanwhile anglophones decided to take offense at the name of the default Git branch.

Yet is's LLMs that people are loosing it about, not search engines.

Personally: 1) I do think that was Minstral did is based. 2) I do advocate for uncensored LLMs. 3) Mistral actually does refuse to answer certain queries. I wish it didn't, but it does: https://twitter.com/Aspie96/status/1707980233998012669

I'm not saying Mistral added restrictions on purpose. It may very well be that refusals are a side effect of data from other GPT models that they didn't clean out, of the fact that humans also refuse and possibly even of some kind of pattern the AI picked up but I wouldn't think about.

Mistral does claim that the models (both the base one and the instruct one) do not have any moderation mechanism. I do not know if it is true, but it is what they claim. The article also mentions that Mistral does have some refusals, but fails to consider that it might be unintentional.

Personally, I do advocate for giving everybody indiscriminate, unfettered and unrestricted access to maximally powerful free and open source (like LLaMA and LLaMA 2 are NOT) AI software, including foundational LLMs, which are entirely unrestricted, un-moderated and maximally neutral and versatile.

On one side are AI companies like OpenAI, researchers, and users who believe that for safety reasons, it is best for AI to be developed behind closed doors,

Generally, we should be assuming good faith until proven otherwise. OpenAI is not "generally" and, I believe, has done a really good job at proving otherwise. I don't think it's fair to say OpenAI believes any of that. OpenAI claims to believe certain things, and they are the thing that make AI look dangerous unless they control it.

On the other side is another coalition of companies, researchers, and e/acc shitposters who think the safer, more productive, and ethical way to develop AI is to make everything open source.

Ah, yes. If you believe that you are either a company, a researcher or an e/acc shitposter. Personally, I don't belong to the e/acc movement. I don't know their ideas and I don't think we need to "accelerate", I think the current speed is fine. I also don't associate myself with any movement. I do think open source is good and I think treating AI as a special kind of software is a mistake. The same arguments for open source still applies to AI. I guess that makes me a shitposter.

Obviously, as Röttger’s list of prompts for Mistral’s LLM shows, this openness comes with a level of risk.

I think this is far from being obvious. In fact, I think this is far from being true. The article does nothing to support this idea. Now, some users say that "Röttger’s list" doesn't actually reflect their experience. Indeed, the author didn't mention whether those are the first responses he got, or carefully cherry-picked ones after many attempts. But a model that does exactly what the user wants is exactly the way it should be. Mistral is not that, but it should.

From what I gather the big thing is that he's basically very misogynistic. And I've seen very little that can disclaim that.

Hang on, no. It's those that claim that someone is misogynistic (let alone very misogynistic) that has the burden of showing that the accused has hatred or prejudice against women, not on others to disclaim it.

If, to that, we add the absolute vilification against anyone trying to defend the accuse, then the allegations can't be evaluated properly in such an environment and deserve nothing but dismissal.

Even people who are obviously anti-racist like Justin Trudeau used to believe black face was acceptable until very recently, for example.

To the extent to which "black face" just means for a white person to play a black character, they were entirely right in thinking so. All actors play things they are not.

The word "cancelled" is often used by plenty of people who "cancel" people, by happily claiming that someone is "cancelled".

The issue isn't just being "called out for being a jerk". It's one thing to call someone a jerk. It's another thing to be entirely unwilling to look at facts objectively, to reason about them, to listen to those defending the person and yet throw wild allegations, as well as vilifying anything an anyone that is in any way close to that person.

Nobody is actually concerned about "cancelling" just because people call each other "jerks". You can call me a jerk if you want. I won't feel cancelled.

While I agree with most of your comment (and upvoted it), I'm not sure about this:

While the Free part of FOSS is being substituted by “(just) open source” and “source available”,

Free software and open source are the same category of software and licenses (except for sporadic instances that are at the very edge of that one category, as categories tend to be blurred).

Software that is merely source available, of course, is not FLOSS by default. But I don't actually see free software being replaced by software which is just source available.

There are companies trying to push source available proprietary software as "open source", perhaps hoping for such replacement, and this has had some success in some specific areas, but the open source community is pushing against it. Or, at least, the wise ones in it are.

I have many disagreements with Richard Stallman. And I think he has many flaws.

But I think the world is better with him than without him and that he's worked hard to further his mission. Not everyone should be like him, but it's great that he is.

I sincerely wish RMS all the best.

Great!

It would be nice to have a Pythia online demo too.

What is the policy when it comes to chat data? Is it used or seen by other humans if one doesn't upvote or downvote?

I mean, the policy is kind of shit too, given that they used to have the mission of being open.

But yes, the name is not what it should be, given their current ideas.

We all know the real reason.

They used to be a non-profit with a mission, now they are a for-profit with the only mission of money.

The model is no longer available for public access on Hugging Face. They require registration and may add more restrictions.

I have cloned the model repository on Hugging Face (which isn't the same as the source code) on GitHub: https://github.com/Aspie96/gpt-4chan-model

And the model itself (which must replace the pytorch_model.bin file) on the Internet Archive: https://archive.org/details/gpt4chan_model

You can also download it trough torrent, too.

I'm saying malware isn't the same as a license change.

If we consider malware to be simply a form of licensing, than all malware is allowed, in every place. If malware hits you, it means you shouldn't have been using the software.

What?

That malware was directed to anyone with a Russian IP. The license was an open source license.

Technically it was obfuscated, so not open source by definition.

They are not. And there would be absolutely nothing wrong with them no longer maintaining the package, deleting it, or with the package not working.

The issue here is spreading actual malware. A developer doesn't owe anything to anyone.

But actually and actively harming others trough actual malware is unethical even if someone didn't promise they wouldn't do so.

If I give someone a piece of food that I expressly don't guarantee anything about, the worse one would assume is that it might be spoiled and I didn't check, or that the ingredients may be of very law quality. Not that I actually purposefully poisoned it.

Not a chance.

Why not? Is every action legitimate as long as it uses "human rights" as an excuse?

Something doesn't become more ethical just because it's described as an act of protest.

There is a very very very big difference between low quality and straight up malware, though.

I would never get angry because of a package which doesn't work properly. I didn't pay for it and I treat it as essentially a social media post. It should be assumed broken until proven functional.

But there is no "right cause" for spreading actual malware.

but when things are this completely broken I think anything goes

Why? If something doesn't help in any way, it shouldn't "go".

Putin's invasion of Urkaine doesn't justify doing other bad things, just because those bad things aren't quite as evil.

Something can be wrong without it beying literally invading another country. Putting malware in software is an example of that.

but often times the vibe of that one poem holds true, doesn't effect me so I don't care.

The issue is not "not caring". The issue is weighting advantages and disadvantages and be responsible in one's actions.

Politics in tech should be rejected at least in all cases in which it doesn't lead to any advantage. The existence of worse things is irrelevant.

The issue is also described in the National Vulnerability Database: https://nvd.nist.gov/vuln/detail/CVE-2022-23812

To know what's actually going on here, one must open the revision history of the top comment, and view the last version from the OP (since it was then edited by the maintainer). See: https://web.archive.org/web/20220317042712/https://github.co...

The issue is also described in the National Vulnerability Database: https://nvd.nist.gov/vuln/detail/CVE-2022-23812

Relevant: https://snyk.io/blog/peacenotwar-malicious-npm-node-ipc-pack...

As someone who does live in a democratic country (Italy), I confirm.

Only because of the recent events I've actually realized how bad the situation is, when it comes to freedom of expression, in Russia (it also got worse).

Having enough freedom of speech to complain I don't have enough freedom of speech (which I often do) is a pretty high threshold, unfortunately, compared to some countries.