I've tried it and so far it's a bit underwhelming. Looks like no model so far can get close to the initial shock I had when chatGPT was released. They're all nerfed to hell and have lots of guardrails.
Well, that makes a lot of sense now :) thanks.
It's a paid service where you can enter a premium link or torrent link to it and it will generate a direct download link. This is very useful if you visit premium sites like Mega and RapidGator where if you don't have an account, it enforces limits such as:
Slow download speed (e.g. max 1MB per second while downloading)
Maximum number of downloads per hour (e.g. 1 file per 5 hours)
No resume support
Unable to download if file is larger than a certain amount (e.g. no more than 5GB per file allowed for non-premium users)
more on the old site: https://old.reddit.com/r/Piracy/comments/q3vqgv/introduction_to_debrid_services/
Is this topic-specific or are there other bots other than the one in UnderNet? I've never found on IRC a book that wasn't in libgen
Didn't they ask why you got banned? What did you do, what did you dooooo?
Reddit answers to shareholders and a board of directors. Each fediverse instance is run by its owner and negotiates with other instances to federate.
If some bigshot wants to own reddit or rule it like a totalitarian asshole, he can buy it and vandalize it if he wants to. If some bigshot wants to own the fediverse, good luck buying the thousands of instances that exist and they still wouldn't get us, because we'd move to or create new instances.
Thanks, I was just testing the bot, but something failed. Thanks for the tldr, yea I guess electrolysis only pays for aluminum, because there is no other way to get it :/ maybe hydrogen too.
Maybe the best option for plastic that ensures it won't end up in nature and is not too expensive is incineration powerplants.
Karma still exists even when you don't see the number, because it is used to sort posts in some way (by number or upvote percentage). Upvotes are important information to the community in principle.
It's not so much a dark pattern, but an emergent property of the upvote system: usually the first commenters tended to have an advantage and late good comments actually would never get enough exposure to float to the top.
Karma farmers would just sit at "new", spam comments and get visibility for joke and outrage comments.
The solution may be to randomly order comments below a certain threshold and/or within an upvote range.
I remember, it was fun. And it's actually important in the age of silomania. Heh, the internet's not dead yet, turns out.
Gsus4
joined 1 year ago
Yea, I'm still using the openAI one and I'm very surprised google can't match it, with the reams of data it collects and has stored over decades. There is something between model design and data quantity that openAI got right and I'm not sure exactly what it was, but I need to find a better name for it than "data cleaning" or "hyperparameter tuning".