monkpit
monkpit on Hacker News
RTO is what you do when you want to make people quit.
> From experience
You’ve experienced all mental illnesses?
Well, maybe if your own media server only showed videos in 1fps compared to hosted ones. You’re not comparing apples to apples here. You might argue that local inference speeds are acceptable, but that doesn’t make them comparably equivalent.
If that worked, we wouldn’t have been in this situation in the first place.
The key is to stop caring (/s?)
It would be cool if there were some way to have a p2p git remote. Not sure how it would work with pushes, but it’s intriguing.
Kind of ruined by all the spam ones that just say like “fs” “sd” etc
… what demos aren’t working?
You’re getting answers, you just don’t like them. It’s faster and cheaper. And if you were to ask an LLM to give you a confidence score it would just be a fabrication.
???
Then why would any AI company exist when they could just use their money to buy tokens from another AI company and make money for zero effort? There would be no incentive to be a provider.
Not to mention inflation would grow to match or outpace the rate you could earn on these guaranteed AI gains.
Agentless passive, hedging, zombie nouns. It’s the type of writing that’s meant to meet a word count.
I think a scooter or a bus is a more apt comparison than a steam engine. The scooter and bus can both get the job done with some acceptable trade-offs, depending on your circumstances and what you’re willing to accept.
However, there are many use cases where they aren’t the right tool for the job.
Yes, I fail to see anything meaningful. If you move the goalposts and say “I have invented a human that cannot be convinced in any way to give a wrong answer” then what’s the point of that in this discussion, really?
And you seem confused about how an LLM works and what it is - “the intelligence of all humanity” - not how it works. You’re debating using 2 imaginary things you created.
What’s your point though, really? “You can train a model to say things that are objectively wrong”? You can do the same with a human.
> You see this already, LLMs are a lot more reliable in statically typed languages with strong memory guarantees (like typescript or rust) than in weaker languages.
It’s common advice to wire in deterministic feedback to your workflow with LLMs - static languages aren’t inherently better for LLMs, it’s that LLMs produce better code when given deterministic feedback, such as compiler results.
Some people report less GI distress with creatine HCL but it’s much more expensive.
The GI issues are caused by the creatine absorbing water. This causes your gut to have more water than usual as it follows the creatine through the works… which causes… urgency.
The effect diminishes after you’ve been taking a regular dose for a while, even with just monogydrate (not HCL).
Some people say HCL causes less distress because it doesn’t draw water as much. Others argue that defeats the purpose of taking it.
Monohydrate is studied to death and proven effective, HCL not so much.
Yeah, I’m missing something here… didn’t he say that it goes down by a page if you click there? I always thought it just took you to that spot.
Apparently the definition of clear is debatable here? It’s ridiculous to think a mechanic would pay people to damage gas tanks in the hopes of drumming up business.
Never attribute to malice… etc.
The point is that it doesn’t matter, they enjoyed it. The fun police weren’t invited to do an investigation, no matter what scenario or analogy is applied.
Saying “you didn’t like it right” is gatekeeping.
The analogy to a book doesn’t fit anyways.
The closest thing would be telling someone they’re reading a choose-your-own-adventure book wrong somehow, and even then it’s a stretch.
Why announce that you’re incapable of allowing someone to hold an opinion that doesn’t match yours?
Would it be so wrong for someone to enjoy something in a different way than you do? Or should they be chastised for enjoying it the wrong way?
You’re welcome to ask what brings them enjoyment, to broaden your views, instead of dismissing them.
To be fair - you’re crafting a deliberately bad model input for a contrived example.
Why can’t someone else have a different opinion than you, even if it doesn’t make sense to you?