There are a wide range of computer skills. Being able to interact with a word processor extremely efficiently is a highly valuable tech skill. Someone who knows about processor architecture but can’t touch type is arguably more tech-savvy but also less useful in most office jobs. So I’d say that the secretaries were indeed tech-savvy in a way that was useful for their positions.
That’s the sad truth of it. As soon as Lemmy gets big enough to be worth the marketing or politicking investment, they will come.
There are a series of quants available at much smaller sizes such that you don’t need the full size model. The full fp16 model is 23 GB but the fp8 quant is half that, and there are GGUF quants down to 6.8 GB.
Oobabooga is a pretty beginner-friendly solution for running LLMs locally. Models are freely available on Huggingface, but look for GGUF quantizations that will fit in your VRAM. The good thing about GGUFs is that they’re typically offered in a wide range of sizes so you can pick one that will fit on your GPU. If you use all your VRAM and start offloading to system memory then the generation will be far slower.
I’ve had the best results with Noromaid20B and Rose20B quants running on a 16GB 4080. Don’t expect it to be as smart as GPT 4.0, but those models do a pretty good job of following instruction and writing decent prose.
Once you mess around with Oobabooga a bit, I’d highly recommend picking up the SillyTavern front-end. Oobabooga runs the actual model while SillyTavern manages characters, world lore, and offers a wide range of other features including a “visual novel” mode where you can set up character sprites that emote based on the content of the messages. It takes a while to get the hang of but it’s pretty cool.
Okay, science guy, but can I use it to replace human workers? I don’t really get all this computer stuff, but I need something to tell the shareholders.
The thing is that I do want to have my subscriptions and favorite channels, and as long as UBO blocks ads, I haven’t fully made the switch to a different front-end.
But it still bothers me that it serves me far-right, religious, and conspiracy theory content given that I’ve never once engaged with any of those topics.
I didn’t read the argument as saying one platform’s behavior excuses the other. I saw it as saying that both are bad.
It certainly doesn’t come across as a defense of either platform to say they’re both infested by Nazis.
I have no idea why you’re being downvoted since you’re 100% correct. I watch one video about gaming and YouTube’s recommendations are all alt-right anti-feminist stuff with Ben Shapiro and Jordan Peterson.
Google surely knows enough about me to know I lean far-left but the algorithm is determined to feed me that slop.
I have no idea from a technical perspective if Odysee’s algorithm is independent from or worse than YouTube’s, but the criticism of YouTube is completely valid.
Neither Meta nor anyone else is hand-curating their dataset. The fact that Facebook is full of grandparents sharing disinformation doesn’t impact what’s in their model.
But all LLMs are going to have accuracy issues because they’re 1) trained on text written by humans who themselves are inaccurate and 2) designed to choose tokens based on probability rather than any internal logic as to whether an answer is factual.
All LLMs are full of shit. That doesn’t mean they’re not fun or even useful in some applications, but you shouldn’t trust anything they write.
It’s a war on both, but especially on LGBTQ people. The fundamentalists are anti-porn in the same way that they are anti-sex in other ways, like opposing sex education.
But it is absolutely part of their strategy to define anything LGBTQ-related as sexual or pornographic, and therefore to criminalize any public visibility of LGBTQ people.
Every metric! Miles per gallon! Head of cattle per acre! Every possible metric!
My trans friends are what keeps me committed to voting Biden no matter how disappointed I am in him. Things are already really scary for them right now and I can’t be complicit in making them worse, even through inaction.
My S10e is cracked but only all over the backplate. Which is also no longer really attached to the phone. I’m driving this thing into the ground.
Bots and grandparents getting fleeced by bots.
But what if you’re not in london
This is what, their fourth?
Only since 2018. Before then I was born on 1/1/1990.
AI can make any blurry criminal look like George Lucas with the right LoRAs.
I said in my original post that just typing a prompt isn’t an example of skill. I stated that there are people who use both AI and non-AI tools in complex workflows that include a ton of manual work, and in those cases it’s disingenuous to write off the process as not being creative.
I’m not sure exactly what you’re arguing against, but it isn’t the position I took. Seems like a reading comprehension issue.
I’d be faster without autocorrect than with. I feel like it chooses the wrong word more often than not.
Honestly, I miss the real keyboard from my 2009 Blackberry. No substitute for haptic feedback.