Just don't leave the country after you participate.

For that matter, you may suffer remote retaliation even if you stay home. I would leave this to the military or other government branch.

Also there is the question of legality, government operation or not. It sounds to me like vigilante justice. And countries who like their anti-American cybercriminals will certainly treat it as illegal. (Hence the advice not to leave the country.)

We certainly need a way to stop cybercrime, international or otherwise. But this doesn't strike me as a good way to go about it.

Finally a prediction: in the not too distant future, one or more offshoots from this will turn into unmanageable domestic criminal organizations.

Just because you have no self-control doesn't mean the rest of us are in need a police state. Even if alcohol was illegal when you were conceived, prohibition showed us that you'd still have been born with FASD. At the same time, if there was no alcohol involved, it's unlikely you'd have been conceived at all.

People do as they will, for good or for worse. What other people do is none of your business. You claim they'll be desperate, yet at the same time you claim that people wearing nice clothes while using a nice car to run over a lady just to steal her Rolex means they did it out of desperation. That's not desperation for the same reason you're not desperate for that second cheeseburger you ate while that skinny hobo next to you had nothing at all. That emotion you're feeling is called greed. You feel that you don't have enough, so you want more. Even though you're already fat, you think it's only fair that somebody with more money than you pays for your third cheeseburger. It's not fair, it's just your greed getting the better of you and making you even fatter.

The whole reason these came about to begin with was WIPO was penalizing the US for not allowing online gambling within its borders. That's part of a treaty obligation, so federal law is going to win here. Nothing NY does can change that, the best they can do is delay it.

But conspiracy theory buzzwords are? Either way, I don't deal in conspiracy theories, nor will they get a dignified response.

I see you've never heard of Anheuser-Busch or Target.

If you read a history book, you'll find:

- The US has always been infighting
- Outsiders who physically attack us suddenly find all of that infighting energy channeled directly into their faces

That's sugar-coating it -- this IS a long-term decline for Hollywood. If this was a pandemic blip, it should have already recovered, just as everything else has done long since then. The film industry itself is fine, just away from Los Angeles.

They probably already wanted to do that anyways. Probably to Atlanta. I've lived in LA for the past 4 years, and pretty much everybody in the movie business is talking about how they can't really get any work here. Main reason why is it's cost prohibitive and overregulated, a lot of both having to do with FilmLA, but a lot of other reasons as well. California is very hostile to its bread and butter industries overall, and I think the government has it in its head that nobody will ever leave. The problem with that *ahem* theorem is many already have. Some have gone to other states, some have gone to Canada and Eastern Europe. See for example, Kobra Kai, which was done almost entirely in Georgia, yet the setting is present-day California. The main thing stopping a major tech exodus is the VC investment is still centered there. That could change after California successfully evicts its resident billionaires, which it seems hell-bent on doing.

One way or another, the problem will self-correct, either in California's favor...or...in California's favor.

1) 30B, dense. Optimized to fit Q4 quantized in a 24GB card with a speculative decoding model as well.

2) Because reporters don't know what weights are and assume you don't know either.

3) Way better than Gemma 4 on text tasks, slightly better on multimodal. Numbers below are all: Benchmark: Muse Glimmer score Gemma 4 31B score difference

Artificial Analysis Intelligence Index: Muse Glimmer: 35 30 +5
MCP Atlas (Public): 75.5 54.2 +21.3
DeepSearch QA: 74.6 61.7 +12.9
SWE-Bench Pro: 51.2 36.9 +14.3
SWE-Bench Verified: 76.0 66.6 +9.4
OSWorld-Verified: 65.9 58.5 +7.4
GAIA2: 43.3 36.4 +6.9
WildClawBench: 47.6 37.6 +10.0
TerminalBench 2.1: 51.7 43.4 +8.3
Tau3-Banking: 23.5 15.1 +8.4
MMMU Pro: 74.0 73.0 +1.0
Charxiv Reasoning: 78.8 77.7 +1.1
OmniDocBench v1.5: 75.8 72.5 +3.3
ScreenSpot Pro: 75.4 75.9 -0.4

4) 128k tokens

5) No, sadly.

What you want is a DGX Spark.

It costs $4000.

And yes, there are fundamental advantages to cloud services, such as large-scale batching, little idle downtime, high speed, and hardware optimized to the specific models / serving needs. That said, one can weigh that off against sovereign control over your server...

Define "tolerable".

The model in question - Muse Glimmer - with DFlash/speculative decoding - will probably get you ~60 to 124 tok/s on a 3090 (a quite dated GPU). For a 5090, it's said to clock in at 233,4 tok/s. On CPU you're looking at maybe 3-5 tok/s.

If you call that "tolerable", I guess you're more patient than me? And as mentioned, you're not just wasting time, but also wasting a lot of power too - CPU is a very power-inefficient way to run ML models.

If you insist on CPU, this isn't the right kind of model anyway. You want to take advantage of the fact that you probably have lots of (comparably cheap) RAM, and compensate for the fact that you have (comparably) terrible memory bandwidth, and for that, you want a MoE with a high total parameters but a low active parameters. Not a dense model like this.

Then I was basically aiming at Mac Minis

That's very much a special case which you didn't mention in your post that I responded to, but still the answer is "meh". You couldn't run it at all on a 16GB Mac Mini, and I think you'd struggle to run it at all on a 24GB (because you have to share the ram with the OS, the inference server, etc). For the base Mini you might get 10-12 tok/s, and for the M2 Pro / M4 Pro, maybe 25-35 tok/s. Still pretty far from a GPU, though.

This model is designed for >= 24GB GPUs.

MoEs don't save RAM (for a given quality), they increase it. You have to store all of the parameters in memory, not just the active params. But inference only uses a subset of the total params for each token, so it reduces the memory bandwidth requirements and improves token generation rate. But this comes at the cost of a higher total param count for a given quality.

Meta has always been releasing open models (Llama was famously the first powerful open model). The change has actually been in the opposite direction, with Muse Spark *not* being open.

Zuck descrbed his motivation way back when, about how they got burned with Facebook on app stores, in that Apple and Google could basically bully them however they wanted, on whatever extractive terms they wanted, and there was nothing Meta could do about it. He's now paranoid about "others controlling the platform", and wanted to make sure that doesn't happen with AI, that they have their own AI base to work with.

I mean, if you want your model to be terribly slow and terribly power inefficient, sure, go ahead and run it on CPU.

This really isn't a good model for CPU. For CPU, you want a MoE with a large number of total params but a tiny number of active params. Something like DeepSeek V4 Flash 0731 if you have at least 128GB of RAM - you might get 2-3 tok/s or so on that. The goal is to minimize the memory bandwidth requirements per token, at the cost of a greater total RAM footprint.

For GPU, you're highly VRAM limited but not bandwidth limited, so your best option is generally a dense model (non-MoE) with speculative decoding to make up for the performance limitations.