I don’t use Google, but DuckDuckGo Search Assist often shows summaries with links that don’t at all support what the summary says, but I often still forget to fact check it. I might start using the no AI version of the site because I’m concerned I might accidentally register something important as factual at some point that isn’t.

Ah, ok, hope it helps someone. I’ll probably try the model this weekend sometime.

This Dockerfile worked for me to build the llama-cpp-turboquant fork: https://huggingface.co/spaces/ai-engineering-at/llama-cpp-turboquant-guide/blob/main/Dockerfile. Should work for upstream too. The Dockerfile I made myself crashed 2 different machines, but then I found this one and can confirm it works well.

From the title, I thought for a moment the weights had leaked. Even if they did, most of us don’t have 10 H100s to run it anyway, so I’ll save my excitement for the upcoming Qwen 3.7 open models, although leaked weights would certainly help open models.

I ain’t givin’ you no tree-fitty, you goddamn Loch Ness Monster! Get your own goddamn money!

This is the digital cage and the reason 3,000+ data centers are being built. I think we are headed towards a historically unprecedented level of surveillance where nobody can do or say anything without the Epstein class being alerted.

I suppose all tech is pretty much like this. Controlled by the people, it’s great. Controlled by the oligarchs, it’s shitty and oppressive.

I can’t decide who’s the shittiest between Sony, Nintendo, and Microsoft. Guess I’ll just stick to PC games and emulators.

Local is potentially even cheaper than that. This guy talks about how to get 17 t/s with a GTX 1060 that has 6GB of VRAM on the Qwen 3.6 35B MoE model: https://m.youtube.com/watch?v=8F_5pdcD3HY. He’s using a fork of llama.cpp with TurboQuant and his newest video made after this one is using an even more optimized 28B version of the model. I have cmake building the llama.cpp fork in a Dockerfile at the moment and we’ll see how this performs on my $800 laptop with a RTX 4060.

I’m also impressed how good OpenCode is compared to Claude Code. Qwen 3.6 is not quite as good as Claude and the MoE version that doesn’t need 24GB+ of VRAM isn’t quite as good as the dense version, but it also doesn’t cost $200 a month with usage limitations and a company training their models on your data. If it’s anywhere near “good enough”, I can see this being a daily driver.

Nice, first I’ve heard of Slate. Privacy-focused EV and looks like they offer customizable trucks and SUVs.

Edit:

Looking at it further, it appears it’s a pick-up truck with an optional SUV conversion kit. I like the tinkerer aspect of it, but the “SUV” would have 2 doors instead of 4, which is certainly not ideal. It’s already a pain in the ass to strap a child into a car seat with 4 doors. I’d also be curious about the passenger safety of those riding in the back seat with regard to how securely the rear roof and frame are attached. I also wonder how this vehicle will get around the U.S. law coming into effect in 2027 where vehicles must have a kill switch. Overall, it seems like a $20k truck would be compelling, but the SUV concept needs more work.

I don’t have an EV, but I can imagine it would be nice to not have to go to the gas station once a week.

I get the feeling the 3,000+ data centers aren’t being built so everyone can generate slop, but instead are going to end up being the infrastructure for the digital cage. AI won’t replace humans, but it can certainly spy on them quite effectively. Big tech and the U.S. government have been trending towards a fascist merger of state and corporate power for quite some time, and the FBI’s recent partnership with Flock and OpenAI becoming an official member of the military industrial complex are not good news. I fear we are headed towards a historically unprecedented ability for governments to track and control the populace. No matter what, these data centers are not for the benefit of the average citizen.

Carbon is the 4th most abundant element in the galaxy. Silicon is twice as rare, so maybe spend 6 month’s salary on a quartz ring instead? Either that, or save up for a down payment on a house. Nah, who needs a place to live when you can have a hunk of mineral, right?

Where I work, Chinese models are banned due to legal concerns. Not just in production, on any company-owned machine. That basically eliminates all the decent open weight models. I’m imagining this type of policy will be more widespread. I suppose it’s because of the potential for legal woes if systems and people are dependent on these models and then federal or state laws impose harsh penalties with little time to react.

I guess they’re running g out of places to pirate training data and they’re desperate because their models aren’t particularly competitive.

Ollama is not open source? It seems to be MIT licensed: https://github.com/ollama/ollama/blob/main/LICENSE. Am I overlooking something?

SoCs like this will definitely be an improvement over power guzzling discrete GPUs. Not sure what kind of crack they’re smoking marketing LLMs as a replacement for a keyboard and mouse. I’m interested in self-hosting LLMs for agentic coding and this looks like a good fit for that, but I won’t even be bothering with OpenClaw, and proprietary Micro$lop is garbage is absolutely out of the question.

When I was a kid, I used to turn over the landscaping timbers around the house to find lizards and such to catch. One time, wasps had built a huge nest under one and I ripped the nest open when I turned it over. Man, did those things sting the ever-loving shit out of me. My whole body was covered in stings.

Any nest I found after that, I made it my mission to get rid of it. I’d put on whatever clothes I had like winter gloves, long sleeve shirts helmets, etc. Of course, whatever I put on always seemed insufficient to protect me and I’d still get the shit stung out of me. What a dumb kid.

If Israel honors the ceasefire, then it’s going to take longer for The Chosen People to own the West Bank, Jordan, Syria, Egypt, Lebanon, etc. All those darkies aren’t going to genocide themselves, you know.