Claude can one shot tasks until you get a larger system then it completely shits itself. These models are nothing more than autocomplete, and they can’t hold large systems in their heads. Anthropic literally tried to rewrite all of bun using Claude, they said they did it and yet it still hasn’t released six months later.
Ah I’m not on 2.6, but I am pretty sure I’m on the test flight. Checking my about shows this though. Mlem 2.5.1 (872)
I’ll try updating.
Ignoring McDonald’s entirely, what chain restaurant doesn’t have meat on the menu?
Not really. The output of LLMs is massively increased global warming, increased prices of goods and services, decreased number of jobs, increased numbers of bugs in software, massive increase in wealth migration from poor to the rich, and increased surveillance. None of that has to do with the quality of the users. Just using it at all results in all of those things.
In what world is a production codebase 75k lines. A single project at my last company was 1.4 million lines of code. Every single one of our microservices had more lines of code than that. And just my team was responsible for 50+ microservices. And thirteen tokens per line is hilariously low.
Holding a week of chat in your head is nothing when 90% of it is stuff that is incorrect or discussed and minds changed. This is exactly why my boss keeps saying conflicting things, the LLM doesn’t know what is true or not. It just regurgitates the random thing from context.
It sure is useful at making my boss post a lot of conflicting things because AI can’t understand context at all. “Force multiplier” doesn’t mean it’s multiplying good things. It can just as easily multiply nonsense.
I’m just gonna hope you’re joking and not that your username is an accurate description.
Knowing when someone is at your doorbell without ringing the doorbell is pretty useful. Same for knowing if your dog got out, or a number of other things. I don’t like sharing my neighbors data with Amazon or the Feds, so running it all locally is perfect. I get to control how long any data is kept for, what it scans for, and none of it leaves my property.
The Internet archive doesn’t have an entire data center just for archiving is what I’m saying. Not that they don’t use data centers.
What on earth am I doing what for?
Which still doesn’t require as much energy as LLMs do. I run exactly that on my own server at home. And my unifi system does it on video all locally with no GPU at all. Video processing for identification is incredibly cheap and easy from a computer processing standpoint these days. You literally can do it on a raspberry pi.
They’re definitely not. Hard drives use a fraction of the energy a single LLM prompt does. A study calculated that every 100 words uses an entire water bottle’s worth of water. That’s a lot of cooling for 100 words. It’s very likely that these companies use several gpus for every single prompt, meaning huge servers for just a few users.
Meanwhile the internet archive has archived a significant portion of the internet with nary a data center in sight.
You can compress video. You can’t compress the energy needed to generate output.
Anyone talking about AI in any context in a normal forum nowadays is discussing LLMs. Nobody cared about machine learning or neural nets that were actually being used for their designed purpose.
I think you would switch your tune if you realized how much data your phone provider is selling about you. SMS is so much worse than Apple’s iMessage, which is e2e encrypted.
Like it’s pretty telling that Apple didn’t give data to the government when a literal terrorist killed a bunch of people, but the Feds can put in a simple request form and get the same data from Google in ten minutes.
The phone providers all sell off your info on their data marketplaces.
I think you’re just sexual.
What does BetterMouse do for you here?
lol what. 1Password literally has one of the best security records out there. They’re the ones actually inventing new more secure ways of doing things. The fact that those are all of the issues over the years and they’re one of the most popular pw managers on the planet is pretty indicative of how strong their security actually is. And they’ve never had a password breach, unlike lastpass.
Seems to me like everything is instead becoming even more bloated with AI stuff. A Google chat tab uses 2 GIGABYTES on my work computer.
Phone: 169 in Safari, inf in Firefox
Computers: uhhhhhhh probably several thousand? It would take quite a while to get a number.
They’re both problems. Why do you think the water usage isn’t a problem?
You can pay after they scan the first item. It won’t go through until they finish the transaction. So put stuff on conveyor. Wait for them to scan first item. Tap your card. Go pack up your bags while they’re scanning. When they’re done you’re done.