Apple’s pretty big on local AI, actually. Their higher-end computers have reasonable GPUs for local AI or other GPU-intensive work.

I use rnote for an “infinite canvas” note taking app, but it doesn’t have much by way of text structuring features like lists, headings, etc. It’s designed for hand-writing.

Not sure the two are comparable. Gaming is a space pretty much entirely occupied by proprietary software. I know there’s FOSS engines and some FOSS gaming-related software. But the norm for that space is far more in the direction of proprietary than any other kind of software I’m aware of.

You’re adding negative connotations to words that have none. That’s like saying that we should rename piracy to “downloading media from alternative sources” or whatever just because corpos are anti-piracy. We use these words because they are precise and concise. Big tech is anti those words. We are pro those words. The word itself is neutral. “Sideloading” is a meaningful category because big tech makes it so by forbidding it, or making users have to use workarounds to do them. “It’s not sideloading; it’s just installing apps onto the phone you own”—well, that would be lovely, but that’s not currently the case.

Yeah it’s nuts. I have a small bladder and just naturally pee a lot. I usually went to the toilet between every lesson at school. Is moving schools an option for your kids?

I’m sure you’re probably right. I haven’t looked into how things like clock speed/FLOPS/etc affect LLM performance.

The main thing is you need ROCm/CUDA/Vulkan support. If you do then just try running some models that fit into your VRAM and see how they perform? On my card, it takes maybe a minute or two to generate an answer, sometimes several minutes if the answer is long. I use small models that are <8GB in size, so not the best open-weight ones, but I don’t use LLMs for anything particularly complex anyway so the dumb ones do the trick for me.

Radeon RX 7600, but yes, you’re right, it’s relative. It’s good enough for any gaming I do, which is not an awful lot. I think it’s considered mid-range for its generation, I guess I was just thinking of the GPUs I see people mention they have which seem much beefier than mine lol.

I’m no AI expert but when I looked into it, it seemed that VRAM was the biggest factor being mentioned in discussions of GPUs, hence why I mentioned it.

Not much, but you can use it for basically anything people use frontier models for, it’ll just be somewhat dumber. I occasionally run LLMs for proof-reading important things that I’ve already hand-written and edited. Sometimes for bouncing off ideas. Sometimes for the kinds of questions I’d normally chuck at a forum but I can get a quicker answer if I ask an LLM. No, I don’t “trust” the answers; anything where I can’t just use my own judgement, the LLM will have given me enough to perform my own research to fact-check it.

I have a lower-end GPU (8GB VRAM) and I can still run smaller models on it with decent performance.

IVPN or AirVPN (depending on whether or not you need port-forwarding) is my VPN recommendation.

Divinity: Original Sin 2. Combat system isn’t perfect (my main criticism is that you can unchain your characters to set your party members up in advantageous positions and then trigger combat, which feels cheaty) but it has some cool mechanics. A lot of elemental interactions, interactions with and creation of different surfaces, factoring in high ground for an advantage, etc. I know BG3 is highly acclaimed also by Larian but I haven’t played it so can’t attest to that combat system.

It would likely have to have a large pool of accounts and randomly pick one per request, to avoid Reddit noticing a too-high volume of requests from one account.

What kind of data is it? If it can be printed, you can just print it and store it somewhere safe, or print multiple copies and distribute it around. Plenty of 80 year old books are perfectly readable. That’d only be books printed in 1946 so nothing crazy. You can pay for professional printing so they can do it on good quality paper and so on.

It’s convenient when there’s no mouse. Moving around on a trackpad can be slow and annoying. But not super necessary for me.

IVPN has been working well for me. Heard AirVPN is good if you need port forwarding too.

Reuters blocks my VPN and the Wayback archived version doesn’t display anything.

Impressive, well done!

No, the comment I responded to was about the ~2014 uptick, not the recent one.

The recent one is obviously AI one way or another. Absolute contribution rate is higher due to AI as well.

They do, but I still use the correct dash in code comments. It’s easy to type with a compose key. I don’t know; I’ve just always done it. I think maybe because English is not my first language so I spent a while learning all the grammatical rules; it feels wrong to deliberately break them to indicate an informal tone, the way a lot of people do online.

If the graph is not just using made up data, then it also measures absolute em-dashes, not em-dash density, so possibly could be an increase in interest in contributing to the kernel?