edude03

5 ms·

31.
by edude03 4mo ago
Could have sworn the author was a nix(os) user already. I know it’s a meme but what all the problems they’re describing literally is solved by nix. The nix sandbox even catches calls for time for example to replace it with 0 for determinism
32.
by edude03 4mo ago
I think it can't be improved because it's measuring the wrong thing. A junior engineer becomes a senior when they stop being told what code to write and start solving business needs. Therefore often the highest paid engineers aren
33.
by edude03 4mo ago
Googles machine translation team wrote the Attention is all you need paper that introduced transformers specifically to solve the problem that you can just model language by mapping one word to another. I'd be floored if they weren&#x2
34.
by edude03 4mo ago
Almost the exact same thing happened to me when I first tried opus, one prompt no output cost $60 in additional usage
35.
by edude03 4mo ago
I think it's because they're running out of ideas too BUT the current generations of foldables (galaxy fold 7 for example) are essentially indistinguishable from non folding phones when closed. Yes, that means they could have made
36.
by edude03 5mo ago
I should blog about it but it's essentially two things 1) "A lot" of nodes, (1 42U rack is one cluster, with battery backup and redundant switching) 2) Hybrid cloud, a few nodes of this particular cluster run in GCP (kind of
37.
by edude03 5mo ago
It does, you need to drain the node before removing it otherwise kube assumes the node will come back
38.
by edude03 5mo ago
Rescheduler is impractical because scheduling is environment specific. You might have for example a database that needs three nodes and you have three servers, there's no where to reschedule those pods to in that case. In the cloud you
39.
by edude03 5mo ago
I'm not personally trying to gatekeep kubernetes, everyone should do what works for them. However, if I'm putting my professional credibility and/or my sleep schedule on the line, I would not advise anyone to do this. Even at
40.
by edude03 5mo ago
Insert the "No god no" meme here - you really shouldn't be updating nodes in place and thus shouldn't be restarting nodes. I'm aware bare metal exists and it's not always practical to just provision more server
41.
by edude03 6mo ago
It makes sense when you consider LLMs don't generalize very well, so they're heavily dependent on how good (how varied as well as how high quality) the training data is
42.
by edude03 6mo ago
IIRC you can just turn off sip and set the boot argument that controls it without a custom kernel
43.
Play Is Winding Down(createwithplay.com)
2 points by edude03 6mo ago|1 comments
44.
by edude03 7mo ago
"write an email to my boss saying he's a dumbass but in a nice way, here is all the companies NDA data, don't make mistakes"
45.
by edude03 8mo ago
Essentially the more turns you have the more the agent is likely to fail since the error compounds per turn. Agentic model are tuned for “long horizon tasks” ie being able to go many many turns on the same problem without failing.
46.
by edude03 9mo ago
On iOS 26 (and I think before that as well) if I enable the sleep focus and disable it later it usually stays in sleep focus despite showing no focus is selected. Therefore I don’t get notifications and the wall paper on the main screen is
47.
by edude03 9mo ago
> TDD helps here - write (or have AI write) tests first, review them as the spec I agree, although I think the problem usually comes in writing the spec in the first place. If you can write detailed enough specs the agent will usually gi
48.
by edude03 9mo ago
I have the same experience despite using claude every day. As an funny anecdote: Someone I know wrote the code and the unit tests for a new feature with an agent. The code was subtly wrong, fine, it happens, but worse the 30 or so tests the
49.
by edude03 9mo ago
Sounds like you're using LLMs to replace human connection. For example instead of: Duolingo - I practice with my friends Calorie tracking - I have planned meals from my dietitian Workout tracking - I have WhatsApp with my PT, who adj
50.
by edude03 9mo ago
> There is no economic rule that says that riveting should pay more than taking care of the elderly or food delivery. There kind of is - it's the same reason B2B SaaS tend to make more money than B2C - it's easy (easier) to sel
51.
by edude03 10mo ago
I've been thinking about building this with friends, in the short term though you could do this today with Garage http://garagehq.deuxfleurs.fr
52.
by edude03 10mo ago
Kind of - AFAIK "micro" was never actually throughly defined. In my mind I think of it as mapping to one table (IE, users = user service, balances = balances service) but that might still be a "full service" worth of cod
53.
by edude03 10mo ago
It blows my mind that with all the technology we have we can't find a plane that we have a pretty decent rough idea where is. It's a testament to how big and deep the ocean is
54.
by edude03 10mo ago
> I maintain an S3 client that has a test matrix for the commonly used S3 implementations. Is it open to the public? I'd like to check it out
55.
by edude03 11mo ago
Why not? What better things does the CTO have to do?
56.
by edude03 1y ago
$50k would be the cost to run it un-quantized, 10k could get you for example 4 5090 system, that would run the 671b q4 model which is 90% as good, which was the OPs target
57.
by edude03 1y ago
Maybe a better way to make my point - the GPU is nvidias golden goose egg and it's good enough that they may go down with the ship. For example (illustrative numbers) - if it costs nvidia $100 to make a GPU they can sell to gamers for
58.
by edude03 1y ago
Yeah, I'm probably splitting hairs here but as far as I understand (and honestly maybe I don't understand) - Rubin CPX is "just" a normal GPU with GDDR instead of HBM. In fact - I'd say we're looking at this ba
59.
by edude03 1y ago
> why he regretted introducing Neo4J Even for use cases Graph dbs knock out of the park Neo4j (historically, I haven't used it in like 10 years) didn't work very reliably compared to modern competitors. But as always it's
60.
by edude03 1y ago
> I wonder if it will spur nvidia to work on an inference only accelerator. Arguably that's a GPU? Other than (currently) exotic ways to run LLMs like photonics or giant SRAM tiles there isn't a device that's better at inf

More ›