FunkyStuff [he/him]

  • 0 Posts
  • 19 Comments
Joined 3 years ago
cake
Cake day: June 9th, 2021

help-circle



  • doubt show me a state in the entire world that doesn’t exist because it has captured a monopoly over legitimate violence. The best the subjects of a state can hope for is that state violence is only ever implicit, but if there was no threat of being put to death or seriously harmed for individuals that threaten the continued existence of a state, that state would cease to be.

    However, it is true that America is particularly brutal with regards to executing civilians. Something that stands out is that, compared to other countries that regularly execute their citizens, there’s a pretty obvious skew in terms of who’s getting the death penalty. Compared to China, for example, the US hasn’t executed anyone for white collar crime in a long time (hopefully someone can find a reference to the last time it happened, I’m not sure where to check) but appears to be killing Black and Muslim folks awfully often. Really makes you think, right?


  • I think there’s an interesting phenomenon where even white normies understand how demonically racist the American institutions are. Ideologically committed racists don’t, but everyone else sees at least part of it. However, because this only gives you a negative assertion (don’t trust what the courts say) and the isn’t really a normative, absolute system we can trust in the absence of any reliable rulings from the hegemonic institutions, we’re just left with a wide space of viable interpretations of reality, which lets people get off the hook for assuming reality must be close-ish to what said racist institutions uphold. That closeness between imagined reality and the reality white supremacy wishes to impose is what allows for people who aren’t ideologically committed racists to passively accept the brutalization and murder of marginalized people. “Oh, I can’t support those cruel acts, but the sad reality is they probably didn’t happen for no reason either” is the refrain of the embarrassed white moderate.







  • No one is telling you you need to do violence, though. No one in Lemmygrad or Hexbear believes that the time to take to the street and start shooting is right now, obviously it wouldn’t achieve anything and would just be a waste of life, that’s why fascists are the ones committing mass shootings. What we believe is that when the contradictions mount to the point that the ruling class clashes down on its opposition with violent force, we need to be organized and ready to carry out the revolution come hell or high water, like every single socialist revolution in the past. Whether or not you personally want to participate is irrelevant because the historical process in which capital undermines its own existence is inevitable, some amount of time from now the crisis of capitalism necessarily must reach a point where its contradictions can’t be reconciled anymore and either the ruling classes succeed at preserving the system, or the workers succeed in transforming it into something new. Furthermore, regardless of your own participation in the violence necessary to maintain capitalism right now, that violence is happening anyway, and it’s orders of magnitude larger than the violence that the left is capable of, even if we were the bloodthirsty maniacs some liberals claim us to be. The black book of communism claims that communism’s death toll nears 100 million, but every 10 years far more than 100 million people die because of preventable illnesses, hunger, and conflict, which are all direct effects of the decaying economic system. If you reject to resist that system, then you’re complicit too, even when so little is asked of you.








  • It won’t be long (maybe 3 years max) before industry adopts some technique for automatically prompting a LLM to generate code to fulfill a certain requirement, then iteratively improve it using test data to get it to pass all test cases. And I’m pretty sure there already are ways to get LLM’s to generate test cases. So this could go nightmarishly wrong very very fast if industry adopts that technology and starts integrating hundreds of unnecessary libraries or pieces of code that the AI just learned to “spam” everywhere so to speak. These things are way dumber than we give them credit for.


  • You have a pretty interesting idea that I hadn’t heard elsewhere. Do you know if there’s been any research to make an AI model learn that way?

    In my own time while I’ve messed around with some ML stuff, I’ve heard of approaches where you try to get the model to accomplish progressively more complex tasks but in the same domain. For example, if you wanted to train a model to control an agent in a physics simulation to walk like a humanoid you’d have it learn to crawl first, like a real human. I guess for an AGI it makes sense that you would have it try to learn a model of the world across different domains like vision, or sound. Heck, since you can plug any kind of input to it you could have it process radio, infrared, whatever else. That way it could have a very complete model of the world.