julianlam

7 ms·

31.
by julianlam 4mo ago
You wouldn't distill a car.
32.
by julianlam 4mo ago
[citation needed]
33.
by julianlam 4mo ago
What do you need a frontier model for, really.
34.
by julianlam 4mo ago
I think the bogeyman would be making possession of a local AI a felony.
35.
by julianlam 4mo ago
A small 4-9B model would be able to run cheaply for this sort of work.
36.
by julianlam 4mo ago
Then we'll amend the policy to instruct LLMs to author PRs as Fred Flintstone, yabba dabba doo!
37.
by julianlam 4mo ago
Very surprised an Apple device can have some atrocious ventilation design. I'm running this model on a Framework 13 and the chassis barely heats up at all while running full tilt.
38.
by julianlam 4mo ago
As a FUTO keyboard user for about a year I'm quite excited for this! Their local voice transcription is top notch and proper swipe gestures would be icing on the cake. Now to fix that issue there tapping "Tuesd" sometimes sug
39.
by julianlam 4mo ago
I think it's interesting that people write off open weight models because they're "a few months behind" proprietary models. I know LLMs move at the speed of light (especially these past few quarters), but if Opus and GPT
40.
by julianlam 4mo ago
I think a particular failing with developers embracing AI is fighting the sunk cost fallacy. While you might not have spent as much time putting together a non-working solution, you still did spend time working with the agent to slap toge
41.
by julianlam 4mo ago
That is a cute sentiment, because I imagine it pales in comparison to the loathing these groups feel towards each other in North America.
42.
by julianlam 4mo ago
Of course. Qwen 3.6 35B-A3B on a Framework 13 with 32GB of memory. Running llama.cpp, 15 tokens per second. Outputs code and text faster than I can parse.
43.
by julianlam 4mo ago
"Memory Wall"? My LLM tells me about those all the time!
44.
by julianlam 4mo ago
6 months is a joke. Companies a thousandth their size are giving free or at-cost access for OSS projects.
45.
by julianlam 4mo ago
s/JS/PHP
46.
by julianlam 4mo ago
Watch the Last Week Tonight segment. Basically, Ticketmaster owns all the concert halls too.
47.
by julianlam 4mo ago
> enough to make it work via profit reinvested from sales. Yes, this is just called running a business.
48.
by julianlam 5mo ago
Are you sure? Prior to today the sub seems to be pretty partial to Qwen.
49.
by julianlam 5mo ago
Last time I tried Gemma 4 (26B-A4B) its memory usage would balloon and consume all of my swap until my machine died. Qwen 3.6 on the other hand barely uses any memory at all for its KV cache.
50.
by julianlam 5mo ago
Respectfully, as an OSS maintainer (not to the scale of nginx or valkey, of course)... if a third-party used an AI agent to rewrite my software in a different language, that gives me absolutely no reason to support that new project. It is i
51.
by julianlam 5mo ago
Model-in-the-middle LLC Checks out, what's the problem? /s
52.
by julianlam 5mo ago
This type of laziness isn't novel. Check out left pad or the two dozen other "utility" packages that could be done in a single line of code.
53.
by julianlam 5mo ago
Does this mean a single pair can be adjusted over time to adapt to a changing prescription? Buying lenses is often the most expensive part, especially for those with astigmatism, second only to the frames themselves, which is another racket
54.
by julianlam 5mo ago
> The reason MCP isn't dead is because practically ~every company on the planet is building an MCP server. I know this because we interact with all of them. Wow if that's not an echo-chamber comment I don't know what is.
55.
by julianlam 5mo ago
I continue to use the model I downloaded... for free?
56.
by julianlam 5mo ago
Why the heck would you want to do this. Using git as your undo chain sounds like a pretty awful thing to do.
57.
by julianlam 5mo ago
Not just you, but uncertain whether it's due to unreviewed slop going to production, or increased demand due to slop generation.
58.
by julianlam 5mo ago
> Taylor has turned his AI chatbot memory off, so that he knows exactly what’s shaping how it responds to him. “The LLM [large language model] doesn’t know who you are, and therefore it won’t bias the results you get,” he says. I found t
59.
by julianlam 5mo ago
As of this month, Qwen3.6 (either the 27B or 35B-A3B) or Gemma 4 are talked about often. Also maybe this will help: https://hnup.date/hn-sota The qwen model is my daily driver this week.
60.
by julianlam 5mo ago
As someone who's built a forum software for 10+ years (NodeBB), I'm glad you found the experience exciting. I find building out forums exceedingly fun too (which is why I've been at it for a decade). Like you, we realized tha

More ›