Compilers are incredibly complex things and the often have undefined behaviors that slip in because the compiler has to interact with the operating system at large, and that means the code actually doesn’t run the same on every platform, or even every version of the same operating system. So, while they are technically deterministic, the reality is that you get surprising and unexpected behaviors in practice.

Also worth nothing that the probability distributions for an LLM are fixed at training time, and that is why the weights are static. When doing inference, models are selecting from a top-k of probable tokens at each autoregressive loop for your typical LLM. When you adjust top-k to 1 or simply set the “temperature” to 0, it will always give you the same result from the same input.

Fun fact, the AI models will continue to be developed for the same reason the US gun problem has never been solved despite the solution being obvious. When

What do our findings say about democracy in America? They certainly constitute troubling news for advocates of “populistic” democracy, who want governments to respond primarily or exclusively to the policy preferences of their citizens. In the United States, our findings indicate, the majority does not rule—at least not in the causal sense of actually determining policy outcomes. When a majority of citizens disagrees with economic elites or with organized interests, they generally lose. Moreover, because of the strong status quo bias built into the U.S. political system, even when fairly large majorities of Americans favor policy change, they generally do not get it.

The reality of the situation is that there is an absolutely insane amount of money being poured into ensuring this tech will continue to be developed. The investors and the companies developing it do not care about public opinion in the slightest. So, as actual empirical research shows, your approach cannot work. But since you’re in a cult you’re unable to understand that.

Meanwhile, these companies are literally telling us what they’re actually afraid of which is open models.

But the elephant in the room here is that the way LLMs are used is just a symptom of the underlying problem. And that’s where the focus needs to be ultimately. However, until that problem can be solved, surrendering these new tools to capitalists would be absurd. There is already a power imbalance, and it would only get worse if we reject technology which the ruling class embraces.

all the undefined behaviors would like to disagree with you there champ

I’ll never recover from this. 🤣

I generally agree, there is a lot of hysteria over this tech, and it will eventually burn itself out. I just find it really disappointing that people on the left refuse to engage in a rational fashion with what’s happening, and what is to be done.

Nope, I’m of the technology is an artifact of the social order crowd, and I understand that attacking symptoms of underlying social relations which are the root causes is the height of idiocy.

It would indeed be pretty stupid to equate new technology with racism. The whole idea of having a moral stance against an inanimate object is the height of idiocy. I have a moral stance against the capitalist system and the way technology is applied within it, but you do you.

The only sect like behavior on here comes from people who think that this tech will just magically go away if they post enough ‘AI slop’ comments on a niche social media forum.

Yet, here you are continuing to talk without saying anything of substance. I also love how in your mind, actually doing something practical about the problem of corps owning AI tools is conservative. Amazing mental gymnastics there champ.

And I don’t know what policy you think that I’m defending that goes against my arguments. But I guess we’ll never know since you won’t actually mention what it is.

My earlier arguments were that we should encourage open model usage to displace closed model usage. The whole problem I’m outlining in my argument is that as long as closed models are the only game in town, that’s what people are going to use. And the way forward is to provide credible alternatives. I even gave you a concrete example of Linux displacing Windows use which you studiously ignored.

Turns out you don’t think my AI stance is in good faith due to your poor reading comprehension skills.

it is rare 🤣

Which is that they don’t understand how statistics and probability work. The LLM generalizes patterns that get imprinted on the neural network during training, and then these patterns are expressed in the given context during inference. The stochastic aspect of the system is what allows the LLM to adopt generalized training to the specific problem being solved. So, the outputs LLMs produced are clearly not random in a sense of playing bingo because randomness is not the dominant property of the system.

have fun

Exactly, that’s the only rational way to look at it. The person submitting the code is responsible for it. Period. How they went about producing it is their own business, but they are the one who takes ownership of the result.

Ah yes, when you’re unable to come up with any actual counter point just say that the detailed explanation of why you’re wrong is not in good faith. Bye troll.

But the incentive structure is only present in the corporate versions of these tools. I’m advocating for building these tools in the open, and we don’t have to mirror what corps are doing. I disagree that these tools are anti-worker at their core, they’re just another form of automation. And we have a concrete example of what happens when this technology is introduced in a rational fashion with China.

Every form of automation will be used to alienate and atomize workers under capitalism, but it’s obviously worse if these tools are also controlled by corporations which will decide who uses these tools and how they use them. While we continue to live under a capitalist regime, open source approach at least helps mitigate the worst ills. On top of that, it directly cuts into the business model of these companies.

Open models are what actually scares them and they state is plainly. We only have to listen to what these companies are saying themselves. They don’t care about boycotts, or some online activism, but they are lobbying to ban open models, and that tells you everything you need to know.

What part of distributed systems are you still struggling with here? Also, models are getting incredibly more efficient every few months. Qwen 3.8 that runs on a laptop today is comparable to Opus 4.6 from around half a year ago which needed a whole data centre.

People working on local models don’t need to compete with whatever the big corps are doing, they just need to make tools that are good enough to use on your machine. Just the same way Linux didn’t have to do everything Windows did to become usable.

And yeah, Luddites had lists of demands just you do, how’s that work out for them?

And if the US somehow did ban open models, I’m sure the vassals would as well, but again who gives a fuck? The whole west is like 13% of the global population. And it’s imploding economically and politically in real time. Who cares what goes on in this genocidal cesspool?

amazing counterpoint as always

That’s what I said?

No, I’m countering what you said by pointing out that there are rational reasons for companies to put models out in the open.

Current techniques cannot give you anything close to the level of inspection you can with software.

So, that just means people need to put the work into improving the techniques. That’s how problems are solved. Last I looked, models like Qwen are actually becoming easier to understand as the article I linked explains.

You always be at a massive disadvantage because of the severally more limited compute available.

Not really because you’d be leveraging a global compute cluster and amortizing the work. But the thing is that it’s not even a competition. Local models just need to be good enough to use. The same way my I don’t need to have a supercomputer at home for my needs, I don’t need a frontier model either. A model like Qwen is already plenty useful, and they will only get better.

raining a model also does nothing for the environmental side of things and just makes things worse

That’s going to continue to happen regardless of how much whinging you do. The question is whether models will be owned by corps or by the public.

The point is to not encourage more people to use the tool that gives control to Anthropic, OpenAI, etc.

You already lost this battle. And on top of that, as you so glibly pointed out in your other reply, an insane amount of money is going into these companies. So these models will obviously continue to be trained and operated. This is the reality:

What do our findings say about democracy in America? They certainly constitute troubling news for advocates of “populistic” democracy, who want governments to respond primarily or exclusively to the policy preferences of their citizens. In the United States, our findings indicate, the majority does not rule—at least not in the causal sense of actually determining policy outcomes. When a majority of citizens disagrees with economic elites or with organized interests, they generally lose. Moreover, because of the strong status quo bias built into the U.S. political system, even when fairly large majorities of Americans favor policy change, they generally do not get it.

Normalizing whinging about AI without doing anything productive just means that people who share none of your concerns will be the ones developing it.