anuramat

7 ms·

121.
by anuramat 6mo ago
been thinking the same, but I imagine you could explicitly separate notes and slop, eg something as simple as cron job that goes through all your notes and creates a PR if there's some easy win: typos, inconsistencies, tags, etc I&#x27
122.
by anuramat 6mo ago
true, but it's the same with humans, we suck at problems with sparse/delayed feedback, which includes science (math would be the exception I guess) sure, humans are obviously better at dealing with it, but the one thing nobody is
123.
by anuramat 6mo ago
> LLMs don't read so it's ok to say "SSD read/write speed", but now that we have something closer to the original meaning of the word, someone always has to point out that "LLMs don't have a soul&quot
124.
by anuramat 6mo ago
they (are supposed to) produce average on average , and the output distribution is (supposed to be) conditioned on the context
125.
by anuramat 6mo ago
weird, for me it was too un-human at first, taking everything literally even if it doesn't make sense; I started being more precise with prompting, to the point where it felt like "metaprogramming in english" claude on the ot
126.
by anuramat 6mo ago
wdym by "prompt and vector is small"? small as in "less tokens"? that should be a positive thing for any kind of estimation in any case, how is this specific to transformers?
127.
by anuramat 6mo ago
whats your setup?
128.
by anuramat 7mo ago
torturing a model with human stupidity probably doesn't align with their position on model welfare ; wondering if they tried bullying it into hacking its way out of the slop gulag
129.
by anuramat 7mo ago
it's not gonna get much more autonomous without self play and major change in architecture
130.
by anuramat 7mo ago
as much as I hate cc, 95% of the issues there are either AI psychosis or user error
131.
by anuramat 7mo ago
I'm not excited because they might be ~lying
132.
by anuramat 7mo ago
imho it was more reasonable back then to claim "agi soon" -- back when nobody really knew how it scales
133.
by anuramat 7mo ago
"some model I don't get to use is much better at benchmarks" pick one or more: comically huge model, test time scaling at 10e12W, benchmark overfit
134.
by anuramat 7mo ago
"oops, our latest unreleased model is so good at hacking, we're afraid of it! literal skynet! more literal than the last time!" almost like they have an incentive to exaggerate
135.
by anuramat 7mo ago
any particular reason you're not using a sandbox?
136.
by anuramat 7mo ago
sure, but here we start with a performance problem, not compute
137.
by anuramat 7mo ago
why would you train a separate model?
138.
by anuramat 7mo ago
otherwise you end up with "get a $20 subscription for 1000% more value -- equivalent to $200 in API usage!!![1]; [1] -- compared to API pricing for american companies on the first weekend of the month between 18:00 and 22:00 UTC+8 duri
139.
by anuramat 7mo ago
novel syntax concepts? like semicolons and lambda functions? > know what Nix will do you would also need a lot of time to understand what exactly happens in uv/cargo/npm/apt under the hood. if this is all that matters, you
140.
by anuramat 7mo ago
from what they wrote, they're just changing how they measure the usage; might even be a good thing if you manage your context right: > This format replaces average per-message estimates for your plan with a direct mapping between to
141.
by anuramat 7mo ago
idgi, shitting on the maintainer takes 10x more time than forking the repo I guess he really needed the latest ci/chore commits
142.
by anuramat 7mo ago
the door wasn't even locked
143.
by anuramat 7mo ago
it's literally just a list
144.
by anuramat 7mo ago
this looks like weeks to you? { pkgs, ...}: { home.packages = with pkgs; [ neovim lazygit ]; }
145.
by anuramat 11mo ago
sometimes they randomly choose the ugliest possible way to do pattern matching, eg multiple blocks of nested "if let" instead of a "match", or a "match" instead of a single "if let" otherwise, works g
146.
by anuramat 11mo ago
I believe ADHD might be neutral/beneficial in some environments, and I also am not a big fan of labels; if it has a non-negative impact on your life -- sure, you don't need the meds but Until I started doing stims, I was regular
147.
by anuramat 11mo ago
Are you taking something for ADHD? I think I had a somewhat similar experience: before antidepressants, I was practically non-verbal; later, I had the typical brain fog/emotional blunting -- still unable to communicate with the therapi
148.
by anuramat 11mo ago
for me, medication was closer to therapy than what I imagined (magic chemical that makes depression disappear) -- a drug changing my thought process (long term and on a deep level, unlike recreational drugs) helped me see myself as somethin
149.
by anuramat 11mo ago
but if they're stacked horizontally, buttons don't fit
150.
by anuramat 11mo ago
I mean, it does take less space (compared to having one per window)

More ›