The purge movies are interesting to me. They came out when I was already and adult, so I missed the cultural moment. If you watch the movies, the plot is patently stupid (if not emotionally gripping) -- so stupid, that I was really surprised to find out how much they resonated with people. Governments don't benefit from violence, and normal everyday people wouldn't want to be violent one day year. Most violent people are regularly violent; they beat their girlfriends for years, and finally kill one of them. They have a long, long rap sheet or an active warrant while they fight the police after being pulled over. etc. Further, general mass violence is extremely destabilizing to governments because most people abhor it, and a government that cannot prevent mass violence is a government with no validity.
The idea that a movie could get so many basic things so badly wrong, that it could float a scenario that made zero sense, but still gain such cultural traction was one of the first times I felt old. It's speaking to a youth with crazy and new ideas. They're ready to make their own grand and terrible mistakes, which a future youth will one day hate them for.
Rally round the flag accounts for "violence done by others against us" and not "random violence against ourselves" so I think it's completely meaningless in the case of this conversation.
With regard to human sacrifice, how much violence have you been involved in? Maybe a few fist fights in middle school or high school? Do you know anyone who has committed real violence? I'll bet you don't. It's exceedingly rare these days compared to historical norms. People might disagree on why its become so rare, but the general rarity is a fact. Even for areas with high violent crime, there is usually a small minority of violent re-offenders, rather than a large majority of people committing random violence.
Does years of domestic violence count? What about shootings I've witnessed but wasn't a part of? There's a school nearby and kids carry guns around here.
> Do you know anyone who has committed real violence?
I don't hang out with the violent types in the first place, so the question makes no sense.
>Does years of domestic violence count? What about shootings I've witnessed but wasn't a part of? There's a school nearby and kids carry guns around here.
No, I mean, per capita, how many people as compared to the number of people you know total, have perpetrated violence.
>I don't hang out with the violent types in the first place, so the question makes no sense.
But if violence were ubiquitous, you wouldn't be able to avoid it. I don't know any violent people either. My point is just that most people (per capita) do not commit violence most of the time. The idea that everyone's just chomping at the bit to perpetrate violence is at odds with reality.
I feel like maybe you missed the point? The concept of the Purge movies is the government is using the Purge as a smoke screen to eliminate the people it considers undesirable. It's supposed to be like Kristallnacht with plausible deniability.
What I think everdrive is saying is that for it to be a smokescreen, there needs to be smoke: if it was only the government participating in the Purge to eliminate undesireables, it'd be extremely obvious and rallied against. But it's not, regular people are engaging with it in droves, which is what they believe to be the unrealistic part.
But most people don't participate, they just hunker down and wait it out. It's a small minority who are aligned with the government who do most of it. The prequel makes this very clear, where the government sent masked mercenaries to participate in the test purge to give them the results they wanted. The movies aren't high art or anything, but their message is pretty clear that they think the average person wouldn't be violent.
You need to consider that what regular people consider normal depends on little more then social consensus. Once you recognize this, the unrealistic part does not sound as unrealistic.
I think calling them ideologues is overly generous for most American politicians. They're vehicles for influence via bribery (which is usually done in a technically legal fashion).
> (which is usually done in a technically legal fashion)
Alas, the Roberts court effectively ruled that it's not really a bribe if it falls short of the third party saying they want the government official to vote/rule a particular way for a particular price, the government official does as asked, and he/she sends the third party a written invoice for services rendered.
Even should that happen, I'm pretty sure Justice Thomas would find the larger offense is that the invoice failed to include sales tax.
Remember our Supreme Court's recent bribery scandal wasn't that a number of justices were accepting gifts but rather that they hasn't disclosed them properly. We're perfectly OK with third parties bribing the judges on our highest court so long as the judges fill out the correct paperwork.
most people's exposure to philosophical concepts is through narrative abstraction, you may be an exception but if you want to interact with laypeople this is inevitably going to be your experience.
The Purge was a terrible movie in so many ways but its message is still extremely resonant. It's also about as subtle as a brick wall, yet I'm afraid that it might have been still a tad too subtle.
Im all for fighting for digital privacy rights. But referencing the purge with a smile begs the question, are you yearning for a purge-like scenario? Some fantastical idea that you’ll be the main character in an anarchic chaos?
Yes, if you pay for {lastname}.name, the full domain is yours, including whatever you want to do with subdomains. However, .name also allowed direct registration of third level domains ({firstname}.{lastname}.name). These were mutually exclusive by domain… so if somebody registers the 3rd level, nobody can directly buy the 2nd level, and vise versa. As far as I understand, it was first come first serve based on the 2nd level domain.
There were also some shenanigans around automatic email routing in that case.
Ok. But why would anyone buy a third level domain when they could have bought the second level? It’s a strange thing.
Were the third level domains cheaper?
Me personally, it wouldn’t even occur to me to register a third level directly, because why bother when you can have the second level and do whatever you want with it.
Buying a third level also came with some email functionality, so if you owned first.last.name, you would be able to forward emails from first@last.name to you. If you have firstlast.name, you need something like first@firstlast.name
When the TLD first opened up you were not allowed to purchase {lastname}.name, it was restricted to third level domains only in an attempt to lure in customers wanting {first name}@{lastname}.name as an email address. But they didn't make it easy to configure that email redirect if you used another registar than verisign, and email has evolved.
Later in a desperate search for MORE MONEY, they opened up 2nd level if no 3rd level already existed under that 2nd. This sounds less like another search for more money with new sales as people are saying, but a reduction in unique domain handling and cost cutting (they can shutdown all those email redirection servers).
I am running on a Ryzen 5950X, 64GB DDR4 Ram and a Radeon RX 6700 XT with 12GB... I don't expect wonders, but I have between 11 and 120 (!) t/s.
I told it an open ended question: "What would you improve in this file at {location} in {bigger project}" and the context window seems to be too small. It's running in circles.
I must be doing something wrong.
Again, I am not expecting wonders and I am totally willing to accept that "I am holding it wrong".
Ideally you want to find something that allows a context window for at least 65k for agent work. Over 100k is better but 32k is about the bare minimum for anything in a harness/agent like Pi. You might have to drop to a lower quantisation to make that possible but context is the key for longer lots of thinking questions like this.
If you drop the thinking level to low or even off the context will go further at the cost of the models intelligence. That might also help you get some answer but since such a question will read so many files into that context its the wrong approach. This is one of the bigger limitations of local LLMs, you can't do work that spans across a lot of files because you are almost always context limited. There is a limit of 1M even with the frontier so even that can't review an entire project. Focussing in on individual files or small groups is more achievable.
So just to be specific the -c needs to be >65536 and depending on the size of what is in the folder might very well not be enough.
Your setup is definitely a monster compared to mine so it should give much better results. I also believe you run dense models with decent results with that setup.
Not sure what llamacpp build you are using but on mine using this fork: https://github.com/TheTom/llama-cpp-turboquant on commit 407f3237bfb3eeaff61546797de3d8c1a96be748 compiled for Pascal, I am able to push my current system to 131k context with minimal loss in quality. For reference this is my config.ini file for this model (Cyber is the better uncensored version btw):
would drop it from fp16 to 8 bits. You don't want to go below 8 bits if you can possibly avoid it and if you really do need to squeeze just drop the v value to q4_0 but leave the key alone. Unlike model quantisation this really makes things worse quickly.
Yeah, you need both a smaller system prompt than OpenCode uses, and a larger context configured in llama.cpp. It's probably running into compaction cycles before it even gets to start thinking about the task at hand.
You'd have better luck with pi instead of OpenCode, and see how far you can get that context window higher. For any halfway actual work you need bare minimum 64k, ideally 128k+. You can use 8bit kv cache, despite whatever folklore you may hear otherwise.
I really and unironically applaud you, when the code you're writing has higher quality than the average AI produces AND it means something.
In the domains I get paid to work in, it simply doesn't matter.
The code was shit to begin with, because of hundreds of hacks due to underspecified or simply wrong requirements, bad code practices, architecture that couldn't keep up but was never fixed due to stubbornness...
Not saying we humans would do better or worse in general, just sharing my experiences.
I have customers where AI usage is absolutely forbidden and others where it's totally fine and colleagues vibecoded mess will come to bite them/us all.
I enjoy programming and producing high quality work. Why would I want to give up the thing I find fun and produce something subpar?
> The code was shit to begin with, because of hundreds of hacks due to underspecified or simply wrong requirements, bad code practices, architecture that couldn't keep up but was never fixed due to stubbornness...
My coworkers have always produced slop, even before LLMs. Unfortunately, as the lead on the project it’s still my responsibility to get that up to a certain quality level or at least make it isolated and malleable enough that it can be changed and we all won’t have a bad time doing it.
I guess what I’m saying is.. I get it. It’s hard to work with other people who are just pushing whatever is given by the LLM. But I still have fun doing it myself.
Some are, others aren't. They just accept it willingly.
Just as in the Purge movies politicians are exempt ;-)
reply