I tried using Claude 3.5 sonnet and … it’s actually not bad. Can someone please come up with a simple logic puzzle that it abysmally fails on so I can feel better? It passed the “nonsense river challenge” and the “how many sisters does the brother have” tests, both of which fooled gpt4.
In this case, the context is definitely humans being born on earth. The entire diatribe I responded to can be summed up as “People have all kinds of ethical and moral objections to surrogacy. In this post, I dismiss all of those without an argument, and instead assign positive moral value to everything that increases the number of lives, including surrogacy.” It’s probably one of the dumbest things I read this week.
AHAHAHAHAAH they had fucking piddly little fans blowing. I hope that made the fire worse.
It gets even worse when you add YC’s claim that it doesn’t “fund ideas” but rather “fund people.” They didn’t find Austen (a shit person) because he had a good idea (he didn’t). They funded Austen (a shit person) because they liked him (a shit person).
Deep into that diatribe:
Some people’s moral intuitions are that nonexistence is preferable to, or not obviously worse than, existence in a less-than-ideal setting. I wholly reject this intuition, and looking at the record of the persistence of life in the face of adversity, belong to a heritage of those who have, time and time again, rejected it. Life is Good.
What a disgustingly privileged thing to say. People have survived in shitty situations so therefore more children in poverty is axiomatically good? This guy deserves poverty. (edit: maybe that’s a bit too far but I fucking hate this guy)
I think you’re being too generous. What they wanted to say is “There are genetic traits associated with intelligence.” However, not inserting probability distributions in every fucking sentence is a class 2 misdemeanor in Rat circles, hence what was written.
This is going to happen in any prediction market whose events are opened and closed manually. Unless you can automatically halt trading when the condition is met (this is probably what people call an “AI-complete” problem), there will always be people who notice the condition is met before the event runner. The unfair trades will certainly be reversed if this is a prediction market worth even a little bit of its salt.
Nah man we had so much fucking dosh flowing in we had no idea what to even do with it! I mean, how could I possibly resist not allocating some of it to my friends?
It really is as simple as that. The dude got a fair bit of attention from his LSTM blog post and got addicted. Turns out, you can’t churn out awesome blog posts that often so you gotta switch to the harder stuff.
so we’re calling “not doing pointless unnecessary work” premature optimization now? cool cool
got told to shut up one too many times. See what happens when you censor people libs?
vexologist here. This certainly is vexing.
did you even experience a single conscious thought while writing that? what fucking potential are you referring to? generating reams of scam messages and Internet spam? automating the only jobs that people actually enjoy doing? seriously, where is the thought?
This all but confirms that all those benchmark evals are in the training set right?
pretty rich of you to call others dumb then post idiotic shit like this.
trumpisn is fascism you dimwit
So we’re just dropping the whole well-regulated militia part of that civil right? They didn’t really mean that part right?
I guess good on you for not having the will to analyze anything beyond the surface level?
Thanks for the suggestions. The LLM is free to use (for now) so I thought I’d poke it and see how much I should actually be paying attention to these things this time around.
Here are its answers. I can’t figure out how to share chats from this god-awful garbage UI so you’ll just have to trust me or try it yourself.
edit: I didn’t do any prompt engineering, just straight copy paste.