It is kinda weird how all the chat type tools all spit out such cheeseball language despite that the raw models almost never generate stuff like that. It honestly just makes me feel like the people making this shit are so caught up in their own hype bubble that it doesn't even register to them as abhorrent.
noosphr 3 hours ago [-]
The chat models are optimised to keep people chatting. The attention economy is still the economy this companies operate in. Which is weird seeing as I'm paying hundreds of dollars per month to get good results.
dan353hehe 3 hours ago [-]
This seems fairly interesting. However I can’t find a link to the original 30 papers?
The link on the repo takes me to some site that wants me to “compile” papers for some reason.
Edit: nvm I found them hidden lower in the read me with a link to a different document.
amelius 3 hours ago [-]
> See IMPLEMENTATION_TRACKS.md for full citations and links
This tagline at the end of the website is one of those formulas that I can't help but link to language models.
I'm not saying it couldn't come from a human but since I started noticing it, I find it so cheesy and patronizing, it gets on my nerves.
And below in the README were the conditions set from the prompt (i.e. "use only NumPy (no deep learning frameworks)")
> Learning is not supposed to be fun. It doesn't have to be actively not fun either, but the primary feeling should be that of effort... - karpathy[1]
And this honestly is low effort. all commits only have hours in between them
[1] https://x.com/karpathy/status/1756380066580455557?lang=en
The link on the repo takes me to some site that wants me to “compile” papers for some reason.
Edit: nvm I found them hidden lower in the read me with a link to a different document.
https://github.com/pageman/sutskever-30-implementations/blob...
https://aman.ai/primers/ai/top-30-papers/