pantsforbirds

7 ms·

61.
by pantsforbirds 2y ago
Zed is written in Rust: https://github.com/zed-industries/zed . It seems relatively straightforward to write an extension for it, but I've never personally tried (I suck at Rust). I will say that Lua is extremely e
62.
by pantsforbirds 2y ago
Fortran is in a weird spot. It's incredible for writing the hardest parts of the programs it's used for, but it honestly sucks for doing the "boring" parts. Doing heavy vectorized workloads in Fortran is nice, but doing
63.
by pantsforbirds 2y ago
There is a reason Fortran has survived so long. Writing matrix/vector heavy code in Fortran is (imo) still more pleasant than C, even when accounting for the idiosyncracies around writing a Fortran program in general. The skill floor i
64.
by pantsforbirds 2y ago
As a ML/AI guy, Python is, imo, the greatest DSL of all time. You get great access to C/C++/Fortran/CUDA code while getting a very high level language for scripting the glue between processes.
65.
by pantsforbirds 2y ago
Matlab also solves a ton of common engineering problems with efficient BLAS wrappers. And it is (or at least used to be) pretty trivial to drop in a CUDA kernel for hotter code paths. It's still great for simulation work imo.
66.
by pantsforbirds 2y ago
This is a really great-looking project. I know I've considered building (a probably worse) version of exactly this on almost every mixed ML + Data Engineering project I've ever worked on. I'm looking forward to testing it out
67.
by pantsforbirds 2y ago
I still love miniconda for DS work. If you want to setup a project to process some videos using some python libraries, you can use conda to install a specific version of ffmpeg into the project without worrying about your system installatio
68.
by pantsforbirds 2y ago
I wonder if DuckDB with parquet storage on S3 (or equivalent) would have been a nice drop-in replacement. Plus DuckDB probably would have done quite well in the ETL pipeline.
69.
by pantsforbirds 2y ago
It's pretty common to run analytics workloads on a fixed schedule. I've setup anlaytics workflows that ingest data every day/week/quarter and generate a report once a day. With workflows like that something like DuckDB c
70.
by pantsforbirds 2y ago
Pricing is actually insane for those benchmark results.
71.
by pantsforbirds 2y ago
This is a very strong offering. I've been really impressed with 3.0 Haiku for smaller tasks, but I'm going to have to test 3.5 Sonnet as our primary pipeline model.
72.
by pantsforbirds 2y ago
It sounded more like they forced him to pick YC or OAI. I don't find that particularly unreasonable either.
73.
by pantsforbirds 2y ago
I assume the desktop app with voice and vision is rolling out to plus users first?
74.
by pantsforbirds 2y ago
> Anduril started by building tech to hunt poor people crossing the southern border Seems pretty inaccurate to describe the usage of surveillance drones as "hunting poor people". I believe drones were a mostly bipartisan approa
75.
by pantsforbirds 2y ago
It is a very practical way to show someone is being quoted though.
76.
by pantsforbirds 2y ago
I wonder if the inspiration for this comes from Bibles? It's a common practice to have direct quotes from Jesus in red.
77.
by pantsforbirds 2y ago
It's wild to see how much hate Anduril gets from the same people who have statements supporting Ukraine in their READMEs or have licenses preventing Russians from using their software. How can you demand support for Ukraine, but then g
78.
by pantsforbirds 2y ago
I think their biggest differentiator is the way they sell their products. They haven't made a habit of winning big research contracts, blowing past the budget, and then blowing past the original per-unit manufacturing price estimates.
79.
by pantsforbirds 2y ago
I think the AI itself is working extremely well. The state-of-the-art LLMs are exceptional at many real-world tasks and are deployed in many production code bases successfully. The more recent AI failures (like the humane pin) are at least
80.
by pantsforbirds 3y ago
If you have any problem that isn't vector math, it's absolutely awful. Even IO is unpleasant
81.
by pantsforbirds 3y ago
It really depends on what you're doing right? Any command line work that I use an LLM to help with is non-destructive. I'll use jq, ffmpeg, etc., but I don't update the initial source.
82.
by pantsforbirds 3y ago
Do you get upset when you see one of create_engine("postgresql://"), create_engine("mysql://"), or create_engine("sqlite://")? The author is just adding a new backend for a common
83.
by pantsforbirds 3y ago
I personally thought the author was pretty clear in the article. My interpretation is that there is no `csvbase-client` API for a user to learn, manage, or even import. The end user continues using their preferred dataframe library, but now
84.
by pantsforbirds 3y ago
How is this different from the risk that a dataframe library will stop supporting S3:// URIs? Why wouldn't Foozle's release notes mention a major protocol change? How is this scenario different from any other library mak
85.
by pantsforbirds 3y ago
LLMs are the best thing to happen to command line tools. I get so much productivity out of them now.
86.
by pantsforbirds 3y ago
I wonder if in < 5 years I can make a game with a local LLM + AI TTS to create realistic NPCs. With enough of these tools I think you could make a very cool world-sim type game.
87.
by pantsforbirds 3y ago
People build in languages they are comfortable with. Golang is fairly easy to ship with and the performance is fine for something like this. Seems like a good tool for the job
88.
by pantsforbirds 3y ago
Arrow has been the most exciting piece of technology I've seen in the last few years. The ecosystem being built around it is amazing, and it's standardizing a bunch of disparate data ecosystems. The arrow ecosystem nets you a grea
89.
by pantsforbirds 3y ago
I think the slow animation + fullscreen notification removes a good portion of the fun of the game. The stats are neat, but I wonder if using a log feed (like a killfeed in call of duty) would be a more enjoyable experience for the user. I
90.
by pantsforbirds 3y ago
Thank you for the source! I'll be sure to give it a thorough read. I understand the (theoretical) uses during a peer/near-peer war, but even outside of the (IMO) poor economic implications, I think it can be extremely detrimental

More ›