davidatbu

8 ms·

1.
by davidatbu 3mo ago
So would you take these claims seriously if they came from OpenAI (since Codex is a pretty lean CLI app)? If so, I think it would be in the spirit of HN to discuss the subject matter of the blogpost (increasingly autonomous coding towards t
2.
by davidatbu 4mo ago
By frontend, do you mean wasm in the browser? I'd have expected multithreading there, especially in wasm, to be extremely unergonomic. Maybe you mean to refer to concurrency?
3.
by davidatbu 4mo ago
Note that something that helped the misinformation was that, on Twitter, there were Kimi employees expressing their surprise that the base model was Kimi K2.5, and their indignation that Cursor didn't credit Kimi. They later deleted th
4.
by davidatbu 4mo ago
Thanks for correcting the title I misremembered. Fwiw, the article did culminate with LSTMs: https://karpathy.github.io/2015/05/21/rnn-effectiveness/ --------------------- EDIT: It looks like you deleted
5.
by davidatbu 4mo ago
I can spare a minute :). This isn't exhaustive because this is just stuff I know of, obviously. - At Stanford, Led research on the first (to my knowledge) crop of joint image/text models. Super widely cited work. - At Tesla, led t
6.
by davidatbu 4mo ago
Thank you!
7.
by davidatbu 4mo ago
Gotcha. I'm genuinely curious: by "impressive", are you referring to coverage? I'd be grateful if you could say a few words about it could be more impressive (e.g, if you indeed meant to talk about coverage, say what fun
8.
by davidatbu 4mo ago
Ah, I just learnt that you don't. Jarred's comment saying exactly that: https://news.ycombinator.com/item?id=48133806
9.
by davidatbu 4mo ago
Do you have some examples?
10.
by davidatbu 4mo ago
Not OP. For this particular use case, I think performance is a primary concern. But if you mean in general, I also totally feel that languages that let you represent more invariants statically are better fit for LLMs. I'd love to see e
11.
by davidatbu 4mo ago
Fwiw, that's not the stated motivation for the rewrite experiment. In fact, the Rust rewrite is slower to compile than the zig code when compiled with their internal fork of zig (tho it is faster when OG zig is used). I don't want
12.
by davidatbu 4mo ago
I'm pretty sure that they have decided that backwards-compat is not the best path for Mojo. Matter of fact, the following is the _last_ item on the roadmap on the home page: > Supporting more of Python's dynamic features like c
13.
by davidatbu 4mo ago
Well, of the top of my head, both chatgpt.com and Gemini have text on their home page to the effect of "AI can make mistakes". I'll bet a few bucks such copy can be found in other places, including the terms of service.
14.
by davidatbu 4mo ago
Do you do rolling deploys?
15.
by davidatbu 5mo ago
I'd love to see a link to these emails, if you have one handy!
16.
by davidatbu 6mo ago
I don't think pypi or npm allow replacing existing packages?
17.
by davidatbu 11mo ago
Are you saying that VSCode runs tsserver in its own NodeJS process? Or are you saying that VSCode uses the NodeJS it ships to run tsserver in a different process?
18.
by davidatbu 11mo ago
Just want to say that as an AI engineer, you and the Latent Space folks are doing work that is extremely useful to me. Without y'all, I'd be forced to doom scroll on X to catch up on the latest developments. I wanted to explicitly
19.
by davidatbu 1y ago
Super interesting! Was this C# or something? is there a write-up/mini-blogpost about this somewhere?
20.
by davidatbu 1y ago
How is enshitification (the gradual degredation of service and products for commercial gain) even related to what's being discussed (the gradual obsoletion of a certain set of skills of an SWE)?
21.
by davidatbu 1y ago
I doubt that Meta (the company that sponsors the work on pyrefly) is looking forward to selling a product based on Python typing (assuming that's what's "what's being glazed in the article").
22.
by davidatbu 1y ago
Koka, Roc-lang come to mind.
23.
by davidatbu 1y ago
IIUC, triton uses Python syntax, but it has a separate compiler (which is kinda what Mojo is doing, except Mojo's syntax is a superset of Python's, instead of a subset, like Triton). I think it's fair to describe it as a dif
24.
by davidatbu 1y ago
Fwiw, Chris has mentioned both of those as lessons he took from Swift that he'd like to avoid for Mojo.
25.
by davidatbu 1y ago
Yeah the rate of progress in AI definitely makes it seem like that from the outside for me too. But having never written cuda, I have to rely on authority to some extent for this question. And it seems to me like few are in a better positio
26.
by davidatbu 1y ago
Good point. But the overall point about Mojo availing a different level of abstraction as compared to Python still stands: I imagine that no amount of magic/operator-fusion/etc in `torch.compile()` would let one get reasonable pe
27.
by davidatbu 1y ago
I assume HLL=Higher Level Language? Mojo definitely avails lower-level facilities than Python. Chris has even described Mojo as "syntactic sugar over MLIR". (For example, the native integer type is defined in library code as a str
28.
by davidatbu 1y ago
So I think a demonstrative example of your claim would be if you knew someone who is as accomplished with regards to compilers, language design, tackling really hard long term projects, but not as good at self promotion, and elaborate on wh
29.
by davidatbu 1y ago
Chris's claim (at least with regards to Triton) is that it avails 80% of the performance, and they're aiming for closer to 100%.
30.
by davidatbu 1y ago
Yeah this is slightly confusing for me as well. Even in this very podcast, being a superset of Python was mentioned as a goal (albeit a long term one).

More ›