You don’t need the OCR step… a lot of LLM models come with “vision” They can transcribe your shaky mobile phone pictures and do their slop thing in one step. You could even wire up an output tool MCP style that feeds the slop back via keyboard USB. The perfect Slop Centipede
it’s an opinionated distro, man. /s
When I had the intense displeasure to work on larger Ruby on Rails projects a few years ago I immediately knew something is off with this guy. His Holier-Than-Though-attitude is not only visible on his blog, but everywhere in the code. Like Omarchy, Rails is also opinionated from top to bottom and things start to go wrong as soon as you leave the DHH approved path. Which basically means, if you use the default configuration (aka Convention over Configuration) everything works fairly well. But as soon as you need to customize anything, things go terribly wrong. For example, DHH hates Javascript and therefore Rails has a grotesquely complicated wrapper that generates Javascript from Rails code. If you have a need to do integrate Javascript beyond that it quickly becomes Trainwreck on Rails.
“Mark Zuckerberg is realizing there’s a limit to ruthless efficiency”.
Cost cutting is not synonymous with efficiency. It’s sad that a business focused magazine doesn’t see through the bullshit.
There is more, but why should I give a fuck about Meta. Let this dumpster fire burn.
Can’t read the whole article bc paywall… But if they are really worried about token cost for converting PDF to a Powerpoint, they ain’t seen nothing yet. Agentic coding the way AI companies push it (multiple agents in parallel with Claude, looping etc) uses way way wayyyyy more tokens than this.
Do they? I only use local models on my GPU and my experience is that Qwen3.6 is so much better than Google’s Gemma 4. I have no comparison to big models, because I refuse to use those. But friends told me that Claude and Co are doing pretty dumb things too while frying the planet
I think a major problem is that it is difficult to prove which IP is in the model data. That’s why the AI companies argue that there isn’t a verbatim copy in the model, and therefore it’s not theft. The law in most countries is not equipped to deal with this scenario
Absolutely true, but that doesn’t change the fact that those AI companies stole the knowledge to train their models and they did this on a massive scale.
It’s so ridiculous to see a guy torrenting a few movies getting jailtime while the AI companies make off with the biggest IP heist in history and get applauded for it.
Good for you, but you and me and most of lemmy are not the target group. Millions of gullible braindead consumerists are.
Kind of weird that Nvidia is in there to boost the plus side. That’s like including a road construction company in a car sales chart
What’s your setup, if I may ask? I’m using llama.cpp router with vscode kilo.ai and qwen3.6-35B-MoE-MTP as a model mostly. It’s surprisingly good as a coding assistant, but I think you have to know what you are doing and know your stuff(aka be an experienced developer) to make it useful. just letting it vibe leads to crap code
deleted by creator
It doesn’t even matter if AI or not. The point is, if you’re too dumb/lazy/busy/whatever to come up with an idea yourself, at least don’t tell your date that you are.
Well they haven’t got much else in their life tbf
Description of the author at the end of the article had me howling. I’ll just leave it here for y’all to enjoy.
Amanda Caswell is the AI Editor at Tom’s Guide and one of today’s leading voices in AI and technology. A celebrated contributor to various news outlets, her sharp insights and relatable storytelling have earned her a loyal readership. Amanda’s work has been recognized with prestigious honors, including outstanding contribution to media. Known for her ability to bring clarity to even the most complex topics, Amanda seamlessly blends innovation and creativity, inspiring readers to embrace the power of AI and emerging technologies. As a certified prompt engineer, she continues to push the boundaries of how humans and AI can work together. Beyond her journalism career, Amanda is a long-distance runner and mom of three. She lives in New Jersey.
Or press Ctrl-S and take a nice walk outside
Just remember that lifetime means Plex’s lifetime, not yours.
I switched from Kubuntu to CachyOS last week, after 10 years or so. CachyOS is based on Arch, and did not disappoint so far, extremely fast, makes Ubuntu look old and sluggish. It’s really impressive. The basic installation was easy. The GUI package manager isn’t as polished but works. A little bit of terminal tweaking was required to install some packages (VMM and KRDC gave me some trouble) but the documentation was ok. Absolutely can recommend.
You could still be identified by a lot of factors and the combination of those. IP address, email if provided, cookies + referrer on clicked links or loaded external images, browser fingerprint, clues from actual content in comments and posts, … It’s not that hard, a whole industry lives on this kind of surveillance data collection.
While this is good advice it only addresses the sender’s perspective of communication. What’s missing is how to deal with people who communicate with malicious intent. You’ll wear yourself out quickly on social media if you don’t learn how to protect yourself against that, and “THINK” is only half of the equation
If you do, at the very least negotiate a much higher salary. They will be desperate, make it hurt them.