As if the cat couldn’t get up there already. It just demanded direct placement.

Reminds me of:

https://www.npr.org/2012/04/13/150566153/airbender-creators-reclaim-their-world-in-korra

As for Korra herself, the show’s creators imagined their headstrong heroine as the kind of girl you might meet on a snowboard. “She’s muscular, and we like that,” Konietzko says. “It’s definitely better than being a waif about to pass out. I know, I look like a waif — who am I to judge?”

Some Nickelodeon executives were worried, says Konietzko, about backing an animated action show with a female lead character. Conventional TV wisdom has it that girls will watch shows about boys, but boys won’t watch shows about girls.

During test screenings, though, boys said they didn’t care that Korra was a girl. They just said she was awesome.

This is proof of why OpenAI’s… opaqueness is so dangerous.

Chat LLMs tend to treat everything like an exam question or essay prompt, as a direct consequence of how the base models are finetuned. The hand is like a pivot point in a physic problem. But more importantly:

• The chat context is sort of their whole world. Again, due to training format. So they tend to stubbornly adhere to what has already been said, and have no real means of self correction.

• While we have no idea what OpenAI actually does, in basically every other open model, the vision component is trained separately from pure text input. Point being these models are alright at the very specific set of vision tasks they’re trained for, but the “coupling” of image input to the bulk of the LLM is very weak. The reasoning they can do over text does not carry over well.

Point I’m trying to make is the biggest lie of Altman is pitching ChatGPT as a general intelligence… It’s not. It’s a dumb, narrow tool, like a drill with specific changeable bits. But they package and market it like it’s “smart”, which is a big fat lie.

Go to any of the smaller AI vendors/models (like Minimax, with a new model today) and they do the opposite of this. They show specific uses in specific harnesses, and hyper optimize for that.

Be careful about assertions like that, though. Misinformation taken as fact is posted all the time… just yesterday, I looked up some picture everyone was harping on in a comment with hundreds of upvotes, and it was totally fake. Probably AI generated. No one even checked.

And when I’ve pointed this out in the past, no one cared. People only prune information if it doesn’t validate their beliefs.

I’m not saying the Factorio dev isn’t a jerk, necessarily, just that I don’t trust Fediverse commenters to have an ounce of information hygiene.

No. You ghost her. And you get everyone you can to ghost her.

Don’t feed the trolls. Anything else is just getting her engagement, even so much as mentioning the issue.

Also:

for some LGBTQ+ people to shoot videos telling how HP helped them to come out of hiding and reveal their true self to the world.

Thousands (millions?) of people have already done this. Hence the HP fanfic community is notorious for heavy LGBTQ+ plots, which is partially why her position is so ridiculous.

I know that Kovarex is a piece of shit but I still enjoy factorio.

You talking about this?

https://nichegamer.com/factorio-founder-kovarex-interview-cancel-culture-and-secret-support/

Eh, that’s problematic, but it feels mild by 2026 standards. Tons of people are abrasive and won’t admit they were wrong or went too far… shrug.

But yeah, I guess I’m saying I agree with you. JK Rowling’s actions are not personality “quirks,” nor one-off events, nor tertiary reactions. Not even close.

Corporate, for now.

Thing is, once they’re out there, they’re free utilities, and they can’t be taken back.

Also, they don’t really need to aggressively scrape the internet. There are many good public datasets now, and the Chinese are already making excellent use of synthetic dataset generation on (relative) shoestring budgets. Also, several nations and other large organizations are already funding open model efforts, but they just haven’t had the opportunity to catch up yet.

That’s pretty much what local ML is.

If open weights LLMs take off, and business users realize they can just finetune tiny specialized models for stuff, OpenAI is toast. All of Big Tech’s bets are. It’s why they keep fanning the “AGI” lie, and why they’re pushing for regulation so hard, why they’re shoving LLMs where they just don’t fit and harping on safety.

To illustrate what I mean more clearly, look at the top comments/replies for the NASA Artemis posts, as an example.

…It’s basically all conspiracy theorists, and government skeptics.

Twitter’s focusing the Artemis posts on them because it’s what they want to see, and most engaging for them.

In the EFF’s case, I’m not just talking about Musk’s influence. The algorithm will only show the EFF to users who would be highly engaged by it. E.g., angry skeptics who wouldn’t be swayed by the EFF anyway, or fans who already agree with the EFF. It’s literally not going to show the EFF to people who need to see it, as Twitter’s metrics would show it as unengaging.

This is the “false image” I keep trying to dispel. Twitter is less and less an “even spread” of exposure like people think it is, like it sort of used to be, more-and-more a hyper focused bubble of what you want to hear, and only what you want to hear. All the changes Musk is making are amplifying that. Maybe that’s fine for some orgs, but there’s no point in the EFF staying in that kind of environment, regardless of ethics.

I feel like the EFF’s messaging is just not going to get through to anyone still on Twitter.

Remember, it’s not a fair forum; it’s an algorithm. And it’s not going to show the EFF to users who need to see it.

So sick of seeing confidently incorrect people opining, using historical examples, when they have never before cracked open a history book and have no idea of the context.

This has always been the case through history.

The issue is Twitter boosts them over less engaging experts. The new problem is the medium. Twitter is not a fair forum, and these takes trend deliberately.

…And I think its really important for scientists (or anyone who believes in science) to recognize that. With all due respect, I do not understand, with everything that’s happened, why they still keep using Twitter.

Even not-fully-reproducible open-weights models are extremely important because they’re poison to OpenAI, and they know it. It makes what they’re trying to commodify and control effectively free and utilitarian.

But there are fully open models, too, with public training data.

It’s anticompetitiveness.

They want to squash open models, and anyone too small to comply with this.

I say this in every thread, but the real AI “battle” is open-weights ML vs OpenAI style tech bro AI. And OpenAI wants precisely no one to realize that.

Not me.

I wanna be there to report every attribution-cropped post I can find, at least in subs where it’s applicable. And repost it without the crop, then tell everyone to watch out for your posts.

Ughhh, I could go on forever, but to keep it short:

• Tech bro enshittification: https://old.reddit.com/r/LocalLLaMA/comments/1p0u8hd/ollamas_enshitification_has_begun_opensource_is/

• Hiding attribution to the actual open source project it’s based on: https://old.reddit.com/r/LocalLLaMA/comments/1jgh0kd/opinion_ollama_is_overhyped_and_its_unethical/

• A huge support drain on llama.cpp, without a single cent, nor a notable contribution, given back.

• Constant bugs and broken models from “quick and dirty” model support updates, just for hype.

• Breaking standard GGUFs.

• Deliberately misnaming models (like the Deepseek Qwen distills and “Deepseek”) for hype.

• Horrible defaults (like ancient default models, 4096 context, really bad/lazy quantizations).

• A bunch of spam, drama, and abuse on Linkedin, Twitter, Reddit and such.

Basically, the devs are Tech Bros. They’re scammer-adjacent. I’ve been in local inference for years, and wouldn’t touch ollama if you paid me to. I’d trust Gemini API over them any day.

I’d recommend base llama.cpp or ik_llama.cpp or kobold.cpp, but if you must use an “turnkey” and popular UI, LMStudio is way better.

But the problem is, if you want a performant local LLM, nothing about local inference is really turnkey. It’s just too hardware sensitive, and moves too fast.

Also, for any interested, desktop inference and quantization is my autistic interest. Ask my anything.

I don’t like Gemma 4 much so far, but if you want to try it anyway:

• On Nvidia with no CPU offloading, watch this PR and run it with TabbyAPI: https://github.com/turboderp-org/exllamav3/pull/185

• With CPU offloading, watch this PR and the mainline llama.cpp issues they link. Once Gemma4 inference isn’t busted, run it in IK or mainline llama.cpp: https://github.com/ikawrakow/ik_llama.cpp/issues/1572

• If you’re on an AMD APU, like a Mini PC server, look at: https://github.com/lemonade-sdk/lemonade

• On an AMD or Intel GPU, either use llama.cpp or kobold.cpp with the vulkan backend.

• Avoid ollama like it’s the plague.

• Learn chat templating and play with it in mikupad before you use a “easy” frontend, so you understand what its doing internally (and know when/how it goes wrong): https://github.com/lmg-anon/mikupad

But TBH I’d point most people to Qwen 3.5/3.6 or Step 3.5 instead. They seem big, but being sparse MoEs, they can run quite quickly on single-GPU desktops: https://huggingface.co/models?other=ik_llama.cpp&sort=modified

There’s a whole lot of interest in locally runnable ML. It was there even before ChatGPT 3.5 started the tech bro hype train, when tinkerers were messing with GPT-J 6B and GAN models.

In a nutshell, it’s basically Lemmy vs Reddit. Local and community-developed vs toxic and corporate.

They seem to have held back the “big” locally runnable model.

It’s also kinda conservative/old, architecture wise: 16-bit weights, sliding window attention interleaved with global attention. No MTP, no QAT (yet), no tightly integrated vision, no hybrid mamba like Qwen/Deepseek, nothing weird like that. It’s especially glaring since we know Google is using an exotic architecture for Gemini, and has basically infinite resources for experimentation.

It also feels kinda “deep fried” like GPT-OSS to me, see: https://github.com/ikawrakow/ik_llama.cpp/issues/1572

it is acting crazy. it can’t do anything without the proper chat template, or it goes crazy.

IMO it’s not very interesting, especially with so many other models that run really well on desktops.

it’s a form of private journalism, private opinion, and private art

But without any of the liability hazard.

This is my issue: the big platforms having their cake and eating it. In one breath, they claim to be little open-platform garage startups that can’t possibly be responsible for the content of their users; they’re just a utility. They need protection from Congress. In another breath, they’re the stewards of generations and children, the only ones responsible enough to tame the internet’s criminality. All while making trillions.

They want to be “private content” protected from the government? Fine. Treat them like it, legally.

It is when it warps the behavior of everyone else around you, and everything in charge of your life.

And I’m not just talking about the lost attention. The algorithms are not neutral.