1·
2 天前Beating a straw man, got it.
- 0 Posts
- 104 Comments
Joined 3 个月前
Cake day: 2026年3月24日
You are not logged in. If you use a Fediverse account that is able to follow users, you can follow this user.
I’m running Qwen on my own hardware.
I haven’t found anything yet that isn’t in its training data that I’d want it to evaluate as a control group, but you’re right that it would be a useful exercise.
Here are some examples of the feedback it has given me:
-
This plot point hasn’t been “earned” and needs more setup to pay off properly
-
This dialog is an exposition dump. Find a better way to show, not tell.
-
This character feels like a vehicle for jokes, and isn’t developed enough.
Most of the advice I’ve gotten so far relates straight back to what I’ve read in writing books and is pretty cut and dry. Some things are a matter of opinion, and I push back when I disagree or when I am deliberately breaking a rule.
Edit:
To your other point, you’re correct that a LLM saying something is good doesn’t mean humans will think so, or vice-versa. A LLM is but one tool in the process, and doesn’t replace real human feedback. For example, with a comedy, do human readers laugh out loud when reading it? A LLM can determine statistically whether something is intended to be a joke and whether the joke is overused, etc., but can’t tell you if the joke is actually funny.
-
I have read a lot of books and do analyze my work in terms of techniques and principles I’ve studied over the years. However, even top professional writers don’t work in a vacuum. TV writers, for example, have “the room” with a team of professional writers, producers, etc. weighing in on all writing decisions. For indies, you don’t have that luxury, and even getting another human who is good at writing to read what you wrote and share detailed feedback is hard, especially when said humans aren’t getting paid to do it full time. Asking friends and family to critique your writing will often result in them trying to spare your feelings, whereas Qwen will happily rip your work to shreds and not care if it just shit all over your passion project.
It’s not my “coach” any more than random people online would be if I posted it in a forum somewhere and no more than a LLM or a human peer reviewing my code is my “coach”. It provides a different perspective to help me see beyond my own biases with feedback I can accept or reject.
Qwen has obviously been trained on writing books and a ton of screenplays. As an experiment, I changed the character names in a classic sitcom script and it was able to identify the series from the writing style and then it also identified the episode. It’s not useful for doing the actual writing, but it does provide useful feedback based on sophisticated statistical analysis of my work compared to its professionally-written training data.
I have code for personal projects that solves problems in novel ways as well as other creative work that I don’t care to let Anthropic and OpenAI train their models on. Is my work worth $14k to them? Well, the value is intangible to me, and I can say at least that companies have paid me a lot more than that for code that took a similar amount of time to write. The major data sources for training LLMs like GitHub, Reddit, Wikipedia, etc. have already been tapped, but they always need more and more data. If you want to give them your data like it’s not worth anything, you do you, but they’re not getting mine. If I need LLMs for personal use, it’s local or nothing.
Your data probably isn’t worth $14k to them in most cases, no. If you’re using it for legitimate purposes, though, still helping them by using their product and sharing your data with them. They’re getting data showing how a real human thinks and interacts with their system, and if you have it help you with any of your own IP that isn’t AI generated, you’re giving them something that is literally priceless.
I started out using GitHub Copilot at work because there was a lot of pressure to use AI, and I was put off by how we were churning through PRs that seemed to work, but having to go back and fix the slop afterwards.
Now I’ve realized that there are skillful ways and unskillful ways to use LLMs, and they can in fact be a useful tool beyond just generating slop. They don’t replace a human thinking critically, but they can automate mundane, routine tasks. They can also summarize text well and suggest options for humans to consider. For example, LLMs reviewing code will often find issues the human reviewers missed.
In addition to coding, I’ve recently been using Qwen locally for screenwriting. It can’t write worth a shit, but it does a good job critiquing my work and pointing out problems with the story structure and the like. For example, I can tell it something like “look at the 7 plot elements described in this MD file and point out where this story does and doesn’t follow this structure”, and the output is quite useful.
While LLMs aren’t the magical silver bullet the tech bros are hyping them up to be, they can still be a useful tool. If they’re just used to generate slop, then no, they’re worse than useless.
Not even worth pirating when there are better open source alternatives.
No, but do you really want them collecting your data? Your prompts and anything else you share is helping them quite a lot.
No, it’s not quite as strong, and especially the initial prefill can take a bit. I also sometimes run into infinite thinking loops where I have to stop it and re-run my last prompt.
It’s surprising how close Qwen 3.6 gets on the benchmarks to Claude models, though. Especially when running locally with 200k context, I’ve found it’s good enough to be a daily driver. Despite the faults, it’s better than paying Anthropic $200 a month so they can rate limit me and collect my data.
That’s the next CEOs problem to solve while the current one is enjoying his golden parachute and sailing around the world. Right now, number is going up!
But how much is the data you’re giving them worth? The other option is don’t give them your money or your data. The Qwen 3.6 MoE model with OpenCode is running pretty well on my RTX 4060 gaming laptop. According the Codacus YouTube channel, it even runs decently in as little as 6GB of VRAM.
Edit:
Fixed typo.
Fair enough. The sources I cited aren’t exactly going above and beyond to prove the claim of 30 people arrested per day by linking to any raw data sources, or even claiming they themselves reviewed the arrest data. I’d like to at least see a day’s worth of examples of people who were arrested and why. It’s quite possible it would mostly be threats of violence and other legitimate reasons for arrests like you stated.
I’ve seen various articles stating 30 arrests a day over social media posts. Here are a few:
https://nypost.com/2025/08/19/world-news/uk-free-speech-struggle-30-arrests-a-day-censorship/
Thousands of people are already getting arrested every year in the UK over social media posts. Good thing the government cares so much about “protecting the children”, because their parents won’t be able to from jail if they say anything the Epstein class doesn’t like, especially now that every post will be conveniently linked to a verified identity. 🙄
Their LLMs might get a little confused now when people ask about British shows on BBC.
Legal won’t allow Chinese models where I work—not just in production, but on employee machines or any company-owned device. I believe the rationale is they don’t want any legal problems if federal or state bans are enacted for “national security”, which certainly isn’t an unfounded fear. I have a feeling a lot of larger companies will be implementing similar policies, and I do also worry that any individual using Chinese models for personal use will be arrested and charged as a terrorist or something. Chinese open weight models like Qwen are fantastic, but it does feel a lot of eggs in one basket.
Most of us here think streaming is useful, despite the fact that streaming services are based on the idea that they own and control all access to media with you renting it from them, raising prices, shoving ads in your face, and removing media from the collection you’re renting from them. Our solution in this community is investing in the open source community around Jellyfin, the Arr stack, etc. so we can still enjoy the benefits of streaming services while owning our own content and not having to financially support companies we don’t agree with. I’d say a lot of us here who are happily using Jellyfin might otherwise have a streaming account if it didn’t exist. Admittedly, I used Netflix until I realized I had better options.
I think the same is true with local LLMs. Not all of us agree that LLMs are useful, but most of us here agree that a few tech giants tightly controlling LLMs and renting them to everyone is not going to be a good thing. Without self-hosted LLMs, many people who do find value using them will go ahead and financially support the rent-seekers who are hell bent on destroying the world for their own financial gain, as well as support them by sharing data that can be used to train their models. Even when you use the free tier of Chat GPT, for example, you’re supporting OpenAI by giving them your prompts that they can use to make their models better.
The ecosystem around running open weight models is rapidly evolving. I’m already running the Qwen 3.6 MoE model with the desktop beta of OpenCode on my gaming laptop and it’s pretty decent. Personally, I’ve found ways to use LLMs where they are actually useful and not just slop generators, though I initially thought they were useless before I spent a lot of time working with them. I’m all for supporting and contributing to this ecosystem so that people can use LLMs without giving their money and data to shithead psychopaths.
Anyone outsourcing critical thinking to LLMs is using them wrong. With software development, for example, quality is speed. Yes, a codebase riddled with technical debt will always collapse under its own weight, regardless of whether humans or LLMs wrote the shitty code, so using LLMs to generate slop that senior engineers aren’t carefully reviewing, or letting LLMs make architectural decisions is the wrong approach. However, skillful use of LLMs can actually yield better quality code than humans alone. Here are some examples:
- LLMs often find issues when reviewing PRs that the human reviewers missed in my experience. Different, focused agents can be used to review code from different perspectives, like security, performance, etc.
- Test coverage can be higher with LLMs automating the test writing. They can often also come up with useful test cases that humans didn’t consider and also make it possible to maintain more varied types of tests that the team otherwise might not take on.
- LLMs are great at fixing lint issues. I’ve seen codebases where lint warnings accumulate because only the errors are getting fixed, but LLMs can fix everything quickly.
- Once solid patterns have been established in a codebase by senior engineers and documented in agent MD files, LLMs can implement features and follow the existing patterns for routine features and bug fixes without getting “creative” and introducing new tools and patterns that complicate things out of boredom.
- LLMs can be used to look at codebases for open source dependencies to better understand how they work in order to utilize them better
- LLMs can suggest alternative libraries, approaches, patterns, etc. with the pros and cons of each. While humans still need to do their own research, this can often be a useful starting point that helps make more informed decisions
No, I meant it was me having a debate with the straw man.