theowaway213456

4 ms·

1.
by theowaway213456 21d ago
Do yourself a favor and look up eBPF - it's incredibly powerful and cool tech. Basically lets you write scripts that run inside the kernel
2.
by theowaway213456 21d ago
At least they clarified they are in a yard rather than a lab.
3.
by theowaway213456 1mo ago
It's the second sentence of the blog post.
4.
by theowaway213456 1mo ago
Site is broken: "Error establishing a database connection"
5.
by theowaway213456 2mo ago
What would it mean for codex to "implement LSPs"
6.
by theowaway213456 2mo ago
errgroup has a SetLimit function to control the max number of concurrent go routines running at once.
7.
by theowaway213456 2mo ago
> Like who really runs frontier LLMs at `max`? Umm what? Lots of people do
8.
by theowaway213456 3mo ago
But how else am I supposed to know when we've reached AGI, until I see an absolutely flawless pelican? All of the pelicans so far have had really weird flaws / quirks so I am always a little interested to see how well these models
9.
by theowaway213456 3mo ago
You can also build your own harness for Claude, right? I think that's not against their ToS? Not saying it's ideal, just saying it's an option
10.
by theowaway213456 3mo ago
This sort of take is so tired and boring, and frankly has zero grounding in reality. "LLMs will never <X>" is constantly being disproven every time they scale up to the next 10X and apply architectural improvements. Their in
11.
by theowaway213456 3mo ago
This seems like more competition for Cerebras? Am I understanding correctly?
12.
by theowaway213456 3mo ago
Obviously the approach you mentioned has the downside of two server round-trips being necessary while the QUERY request only requires a single round-trip. Not to mention the two-request approach adds more complexity to both clients and serv
13.
by theowaway213456 4mo ago
Every time you write a test that handles some data, you write an assertion about how much data is handled? Come on, this is such an easy thing to forget to test. Don't act like there is some magical testing strategy that would have cau
14.
by theowaway213456 4mo ago
TL;DR don't have your agent write skills using only its latent knowledge, otherwise you may as well not use a skill in the first place and let it summon that latent knowledge on the fly. Not sure if this take is correct though. I suspe
15.
by theowaway213456 4mo ago
Claiming that something is someone's "entire personality" is extremely reductive and definitely hostile
16.
by theowaway213456 4mo ago
Yep that's the first time I noticed it. The pocket was also numbered for me as well
17.
by theowaway213456 4mo ago
Not a single mention of prompt caching in this article, which is a massive benefit of append-only context.
18.
by theowaway213456 4mo ago
Certain entities seem to be displayed as numbers for me, like "You received a 6" etc when getting my first potion. Anyone else seeing this bug?
19.
by theowaway213456 5mo ago
This implementation is better than stdlib's implementation in my opinion, since it respects context: https://github.com/janos/singleflight
20.
by theowaway213456 5mo ago
Old news - both of these vulnerabilities made top headlines on HN. This post isn't contributing much. (Tangentially, the title of this article is egregiously difficult to parse.)
21.
by theowaway213456 5mo ago
It's very hard to use on mobile. When clicking a graph node, something happens kind of far down in the viewport, which I didn't notice at first. Also clicking "Open post" doesn't seem to work
22.
by theowaway213456 5mo ago
This feels like a suboptimal solution to me because I personally like to keep comments in "unresolved" state so that they remain visible and other folks can weigh in on them if they want, but in a way that doesn't block the P
23.
by theowaway213456 6mo ago
Yup. No matter how much you tell it to keep things simple, modular, crisp, whatever, it generates tons of garbage much too often.
24.
by theowaway213456 6mo ago
lol after reading your first sentence I literally thought to myself "this sounds like the type of person who never closes their browser tabs"
25.
by theowaway213456 6mo ago
In my experience, this mistake happens all the time for native English speakers born in the US.
26.
by theowaway213456 6mo ago
If we're allowed to make modifications here then it should really be lose => looze and loose => luce
27.
by theowaway213456 6mo ago
I 100% understand why you are using a subscription-based model. It makes sense, and I agree it's the most honest model given that you have to continually support it and you don't want to have to either over-promise on extended sup
28.
by theowaway213456 6mo ago
Sorry, my attention started drifting when you said "epistolary."
29.
by theowaway213456 6mo ago
In other words, it requires a tremendous amount of effort to fully communicate your tastes to the AI. Not everybody wants to expend the time or mental effort doing this! (Once we have more direct brain/computer interfaces, this effort
30.
by theowaway213456 6mo ago
Feels like AI generated market research. There must be a single company/person putting out these posts because I see a ton of posts like this following a similar structure

More ›