anuramat

5 ms·

1.
by anuramat 1mo ago
do you think typing `make` into a terminal is some sort of an insane intellectual achievement?
2.
by anuramat 2mo ago
why can't they just increase the price for cached input instead?
3.
by anuramat 2mo ago
could it be that the leading AI lab is decent at making LLMs?
4.
by anuramat 2mo ago
what's there to hack that you couldn't do with $3 in openrouter credits?
5.
by anuramat 2mo ago
> productivity findings are you referring to the early 2025 METR study?
6.
by anuramat 2mo ago
> you can run a small model but why?
7.
by anuramat 2mo ago
I think you mean mostly-stateless-but-with-prompt-caching-and-batching, ie not really
8.
by anuramat 2mo ago
you think they're doing inference at a loss even with the API prices?
9.
by anuramat 2mo ago
yes; fyi usage limits on the $200 claude sub correspond to at least $1.2k/week in api tokens
10.
by anuramat 2mo ago
"benchmaxxing by generalizing" is not really benchmaxxing
11.
by anuramat 2mo ago
> Russia lmao
12.
by anuramat 2mo ago
> possibility of "LLMs can reason not like a human" what would be the difference between not reasoning and reasoning not like a human?
13.
by anuramat 2mo ago
I keep asking the same question, and I think the steelman version would be "has metacognitive patterns similar to humans"
14.
by anuramat 2mo ago
I'm somewhat serious -- if you think AI will scale that well, you can't really make predictions like that I personally don't think the weight efficiency will improve that much; if anything big does happen, I expect it to be a
15.
by anuramat 2mo ago
just found a decent looking benchmark for iterative development: https://swe-milestone.com/ surprised it isn't a bigger thing, eg artificial analysis doesn't report anything like that still doesn't measure th
16.
by anuramat 2mo ago
> nowhere near readteaming I think any security-related task triggers it to think about the threat model and thus hit the guardrails
17.
by anuramat 2mo ago
what languages are you working with? I imagine if it's something like C, you'd hit guardrails every time you manage memory
18.
by anuramat 2mo ago
is there a way to use it with lazygit? unfortunately I'm addicted atp, and "autorebase branches" is exactly what I was missing the entire time
19.
by anuramat 2mo ago
why would you need a local fable at that point? AGI will surely solve all the problems in the world at that point
20.
by anuramat 2mo ago
what are you working on? I only hit the guardrails twice after burning through two weeks of 20x max plan, both times on ML stuff; still more than I'd want to, but not unusable
21.
by anuramat 2mo ago
ml research, both brainstorming and running experiments eg I'd throw a hypothesis at it in the evening, and overnight it would write the code, do a sanity check, start a run, monitor the metrics, identify and fix a bug, propose a new h
22.
by anuramat 2mo ago
security researchers are using the free tokens that they get to do useful security work, AI labs are giving away free tokens to maximize their profits; is it really that hard to imagine that different parties might have different goals? &gt
23.
by anuramat 2mo ago
so, git?
24.
by anuramat 2mo ago
> it's not actually a financially efficient way ... unless you profit from creating demand for LLMs well, they do? it's a win-win, you can't really criticise an AI lab for doing AI instead of straight up giving money to se
25.
by anuramat 2mo ago
I imagine one could one-shot a basic app and then feed feature requests one by one, sounds like an obvious way to benchmark architecture/maintainability
26.
by anuramat 2mo ago
> data centers being built in their communities > golf courses are a traditional green space where people in a community I have a feeling those two sets of communities are disjoint
27.
by anuramat 2mo ago
if you don't make your own tech, standards will be defined by people that do
28.
by anuramat 2mo ago
whats the point of a commit message if it can be inferred by an agent from the diff?
29.
by anuramat 2mo ago
why?
30.
by anuramat 2mo ago
the article implies that the perfect balance is 1:1; I want to believe this is just some sort of a ragebait-based PR strategy

More ›