bioemerl

The Russians have interests in holding a security border around themselves which extends to about a dozen American allies and trading partners Russia extends their security border to those allies and trading partners, they will almost certainly stop being so.

The Russian plans don't stop at Ukraine, they extended to Moldova, Poland, and lots of other nations.

Is it viable for them to do that in the long term? You never know. Right now it seems clear that the answer is no, they can barely take Ukraine so how are they going to handle Poland and the other countries when they've already expended their economic tools that could have done so?

But it's generally in our better interest so make sure that they are as incapable as possible of expanding those borders.

We would be fools to allow them to expand those security borders because it means losing those allies and a huge amount of trade and resources that come with them.

They are either in our circle, or they are in Russia's. Russia treats these countries as former colonies, so as long as they are given freedom of choice they're going to ally themselves strongly with the west.

For the relatively low price of funding Ukraine we slow down Russia's advance at worst, and we totally halt it at best. If we manage to totally halt Russian advances and equip Ukraine to be competitors for Russia, we've now got a country that has just survived a war with Russia sitting on their borders armed to the teeth and angry at them.

They are literally fighting the war for us, without risking nuclear conflict. Why shouldn't we be interested in this tragedy?

> You always have to take responsibility for your actions, and Ukrainians pay for these now.

We aren't defending Ukraine because we are nice people, or defending them because we quite like the consequences of Ukraine's decision, and we're more than happy to support and ensure that those consequences are as minimal as possible, while ensuring the consequences of Russia's decision are as high as possible.

To be fair, might absolutely does make right in international politics still. The only reason we are even talking about Ukraine is that the Russian 3-day invasion totally failed and open up the opportunity for the west to sanction them and support Ukraine.

It weren't for that fact, the ability for countries like Germany to trade oil with Russia still would have totally overridden any possibility of sanctions.

And the West very frequently does not get involved in areas where terrible things are happening so long as those areas aren't really in their interests.

Which doesn't justify Russia in any form shape or way. Personally, because it's individuals we aren't involved in international politics and we are still perfectly valid in criticizing Russia for being hilariously evil.

And because in the world of might makes right, Russia is also losing.

So it doesn't really matter. Any Russian who tries to make this argument might be correct in calling out hypocrisy, but their weak and stagnant nation lacks the might to make the right.

Because Ukraine is a democratic state/western ally being invaded by it's much bigger and meaner neighbor and will cease to exist if we do not help them defend themselves.

We don't have to give them to Ukraine, but doing so generally furthers our interests.

I've been learning it thanks to this amazing tutorial series by a guy named mangojelly.

https://www.youtube.com/playlist?list=PLWuyJLVUNtc0UszswD0oD...

freeCAD had a critical issue in topological naming (they are working on a fix as we speak and it's going to be out "soon"), and it's hideously clunky. Basically, when you resize a sketch and that causes more faces to be created it makes all your other stuff go crazy.

You can also use realthunders branch which fixes the naming issue as well.

Mangos videos teach you a bit how to think reasonably though freecads infuriating errors, too. The guy is worth a watch if you want to start using the best open source tool.

freeCAD sucks, until you learn it, and then it works alright. Learn to use data planes instead of sketch on face and the topo naming isn't even an issue anymore.

We need them to give to Ukraine, no? I would suggest they are preventing harm in that case.

It's certainly a consequence of interest rates going up. In the words of buffet, Reddit has no pants.

They should have improved their platform and lock useful features behind a paywall, like video and image hosting. They could have made money on their core appeal - hosting communities for people.

Instead, they've bloated their site into an adware riddled clusterfuck, ignored useful features that people actually wanted, and just turned into a hilariously useless company in general, making their site worse with every change instead of better.

We desperately need regulation that kills the network effect that feeds these companies, because it's clear their users aren't the priority.

Yes and no.

Yes, it actually works. That's what matters at the end of the day.

But no, eventually you're going to have to fine tune on something that has a larger context or training a brand new model with the position embeddings semi randomized so that it learns to generalize in the first place instead of needing hacks like this to function.

But training a model costs millions and millions of dollars, so we're going to have to wait until some very generous group decides to do all that training and then release that model open source.

Or releases the model file for a paid charge fee, I'd pay like $200 for a good one.

It just keeps getting better!

This is so exciting because contacts length has been a problem with these models for so long and it's awesome to see open source finally cracking that egg.

I need another GPU.

edit:

I'm actually a little bit skeptical of this. Yes it's dynamically scaling which is great when you have a model that's not fine tuned, but I think it's not going to work out too well when you try fine tune a model on a target that's moving like that. I'd rather one that stay static so that perplexity is always increasing up to the max rather than doing much those graph shows where it gets worse over time.

That said I don't really know what I'm talking about so maybe it'll be better regardless.

What if the place with excess energy is America in the place was not enough is Europe?

Iron is one of the most abundant elements on earth, but also we aren't going to be burning it like we do fuels now. Instead we would produce pure iron using green power and oxidize it back to it's natural state for energy in a cycle.

Imagine loading an iron rod into your car, driving for a while, and then when you get into the gas station you dump a pile of rust off and buy a fresh iron bar.

Not really. If your intent is to truly ensure content is labeled open AI isn't able to ensure that since you can just copy paste their output.

They want the law and they want to promise safety while not being impacted by the overbearing regulation they've invited onto the rest of us.

They submitted a paper to congress talking about how they should regulate AI.

HN is not wrong about the fact that there were trying to push regulation in the United States that would help their business. It just turns out the European laws will actually hurt their business.

The two are roughly the same.

A big truck is more deadly to smaller vehicles and we shouldn't make it unsafe to drive a regular sized car instead of one of these monstrously large trucks.

You could regulate things like requiring stop gates on trucks like semi trucks have though, which would at least help.

I think you're mistaken at least with current AI. It's very easy to condition them into a new frame or mind if you speak to them in a certain way. They're also very prone to "trope" - where when your sprinkle a few hints of something that's common they'll fall right into line with behavior associated to it.

We forget that AI becomes more human as it gets more intelligent. These are not computers programmed with a hard limitation anymore, they're actually more prone to being manipulated than other humans are.

I'm spoiled by 4 bit and unfortunately it doesn't appear to be supposed here so this isn't of much use to me, but it's awesome to see people working on the inference speed side of things regardless.

> It’s too easy to be independent these days,

To be fair, I didn't say you have to have a good common cause. Just something to do in some capacity.

Like, you don't need help, it's fun to hang out and it's an excuse to get help.

My point is to say the relationships are different rather than transactional. Focused on a common cause and that sharing of it rather than relationships directly.

I honestly don't think it's as big a problem as you would expect. Guys aren't quite so built to be huggy friendly and we tend to be more defined by real working relationships.

Or maybe it's just a me thing.

Like, at work for me I would be most rewarded by other people who are engaged in that common work cause and interact within that context. Not with stories about family life or opening up about whatever.

Hanging out with friends is kind of pointless without a common cause. To do play some game or sport or actually do something that's worth doing and is fairly collaborative.

To me a friend is like "hey I'm building a fence, wanna help?". And the ability to be building a fence yourself and asking the same. Or whatever project you can imagine.

The "you good" shit is mostly a consequence of lacking that sort of common cause, at least in my experience.

Or just watching other people do stuff is good too. "Hold my beer" silly stuff that you can watch and joke and one up each other about.

But if someone started that family huggy stuff (barring needing help or talking about their problems) I'd be distancing myself from that hard. In my experience those people are out to get you, smile on a snake.

Imports through the browser do, but I'm wanting a single file bundle with all those imports so I'm not relying on a browser feature to deliver the content over N requests.

I know that's less of a concern nowadays with modern http, but guess who still supports ie11

Instance of doesn't work sometimes and you have to write explicit property checks. I forget the details because I'm by no means a typescript expert, but I remember being very frustrated by the fact I couldn't just use instance of.

Honestly typescript is fine, just give me:

Standard library that isn't worthless - no more npm package spam.

Type knowledge in the runtime. It doesn't have to be typed, just make it so I don't have to do crazy checks on properties before casting to a class.

Imports without webpack.

I do not think of my argument rests on the degree to which I am an LLM. I was fairly specific in speaking about the ways in which an LLM is different and the ways in which it lacks the functional capacity to be aware of itself.

I made a mistake here in in using the word sentient the way I did. To help you understand here, I use these words as follows.

Sentient - to be aware of yourself.

Sapient - to be humanlike in your ability to think and speak

Conscious - to have a self experience of any form.

What I am saying is that an LLM is definitely not aware of itself. It doesn't receive itself as an input, so it cannot have thoughts like "I'm happy right now". It's functionally incapable of doing so because it has to receive itself as an input in order to do that.

However, it's possible that such a machine still has a experience of some form in the sense that there is some ethereal consciousness created by the fact that exists and is thinking, but we can't observe that aspect of the world so it's not really something we can discuss in a functional way beyond speculation.

I would argue that the silicon-based LLMs we have today do explicitly lack sentience.

Sentence, self-awareness, can only exist when you receive yourself as an input in some form shape or way.

These llms just don't do that. Technically recursive networks that we were using before the llms could, but these new models have no method but the context to understand their own state of mind.

And my knowledge they aren't currently trained on their own output when it comes to context, they're just trained on a data set that exists already. As a result they have no way of differentiating their own output from the input they get from the world. I don't see them being able to be sentient in any way as a result.

Which is not me saying they can't be, we just have to get to the point that we are training that into them somehow.

For me this is most significant in the sense that I can use it at my workplace.

No, don't tell me that, I'm going to need more graphics cards now.

> This is, I think, the problem. You're prioritizing your fear of an external threat over what I see as the much more real concern of growing democratic censorship within the US

No, both things are issues and both should be concerns. I generally support open source platforms because they generally do allow for people to speak their mind more freely and federation allows for a more open system.

If I was fearful of opinions from communists or people with bad political opinions I wouldn't support decentralized software at all. I'd support centralized clothes communities like reddit that have "the right opinions".

This isn't the case of being single-minded, this is a case of just genuinely the external threat being real and you needing to respond to it.

This is where I come back to pointing that your arguing in bad faith here.

You're trying to argue to absurdity.

Here I have a concrete example of a group that is explicitly politically extreme. They are explicitly in contrast to the values of open source.

Linux, meanwhile, is an absolutely massive project with billions of eyes on it every day which is head and resisted multiple attempts to compromise it.

I'm confident that the software I'm running still has bugs holes and inroads to malicious actors.

But, like you said, you basically can't use a computer if you want to avoid them all together.

So you avoid the obvious cases, you do what you can, and you keep on going forward.

Same story with buying stuff made in China. You don't do it when you can, you do it when you have to. You support laws and regulations that encourage people to move away from it.

Be pragmatic, focus on what you can accomplish, and don't worry about being perfect.

> The software is open source, so if the creators put crap into it for censorship they will rightly get called out for it

Unless they sneak something into a binary.

Or create a backdoor that doesn't get noticed for months or years.

Or they just design the software to make it difficult to weed out incursion attempts from butt networks like we've seen Russia and China use to influence American opinions.

Or they use the fact that basically everyone joins the main instance of these networks to run regular censorship and defederate with anyone who goes against that.

There are many many have a news for abuse. We're only assume that because something is open source that it is trustworthy. When you make that assumption you're making a bad assumption and you're making yourself not secure.

Security takes actual vigilance, and a lot of that vigilance comes in the form of trusting the person you're using the code of.

> When people get the sense that their expressed political opinions can have lasting negative effects on their lives and projects, they become much more reticent

At the end of the day my priority of not enabling Chinese control over our lives in society is far more significant than protecting the Lemmy developers feelings.

I agree with you in abstract, we should hesitate to enact things like this unless it is strictly very important to do so.

This is a case where it is strictly very important to do so.

I'm making the argument that because the developers of living support China and their forms of authoritarian control they are likely to publish updates and make changes to their software which enables and helps to promote those ideals that are ultimately against the ideals of all things open source.

And at the end of the day unless you want to be running software on the public internet which is out of date, you're going to be running their updates, which gives them long-term control and ability to mess with the network after time.

Plus the basic fact that writing such a large and popularly used piece of software gives you a fair amount of social credibility and attention which is not served by those people.

If someone is willing to go and do the legwork of doing a full audit and commit to maintenance of the project in the future I would totally run that code.

Until someone does, you should stay the heck away from it.

Because it's obvious and blatant?

I wouldn't want them to be hidden either, but I'm not going to notice if it's hidden and then I would tell you the same thing except instead of obvious and blatant I would say sneaky weasly and malicious.

I detailed very specifically why their opinions are of interest here.

Unless you have a reason to tell me that their support for a country which actively and happily censors people and the ideals which allow that censorship to happen is not relevant, you aren't making a solid point here.

This isn't just a cute little disagreement of opinion that you can brush over.

And you cannot disconnect the developers from the software. They are very connected in the developers still have a lot of influence over it, even if it's open source.