jwr

4 ms·

1.
by jwr 3d ago
Complaining on HN :-) More seriously, I have not found a solution.
2.
by jwr 5d ago
Every ad system I tried resulted in nearly 100% bot traffic to my site. Google and LinkedIn give you "tools" that will show you various numbers, but I implemented my own conversion tracking, because I didn't trust their numbe
3.
by jwr 7d ago
I find it rather sad that the original posts have been posted on an open site, where anyone can read them, while the HN post points to an increasingly dubious site that doesn't even show the post unless you create an account and log in
4.
by jwr 8d ago
I always thought it was enough to switch off the "Improve the model for everyone" setting on chatgpt.com: "Allow your content to be used to train our models, which makes ChatGPT better for you and everyone who uses it. We tak
5.
by jwr 8d ago
I also have a lifetime pass, but I'm becoming increasingly annoyed with Plex. Why do I have to regularly turn off their "recommended" stuff that is just ads for movies I can buy online? Why does it appear and re-appear? Also,
6.
by jwr 9d ago
I have an LG TV. It has been pestering me with those licensing agreements from time to time, but interestingly enough, I was only required to agree to fairly reasonable basics, everything else I just ignored and left for later. I'm g
7.
by jwr 10d ago
I gave up on reporting spam. Almost all of the spam I get (and I get a lot of it!) comes from Google, Amazon and Microsoft. None of them care much, none of them make reporting easy, and it seems they don't follow up, because the flood
8.
by jwr 11d ago
Acceptance rate is good, but MTP doesn't help in my case because of my machine's memory bandwidth constraints (M4 Max). Turn out it's better to turn MTP off.
9.
by jwr 11d ago
I would (tentatively) agree — but that doesn't invalidate my argument: qualia are not objectively measurable, so you still can't tell if a machine is conscious.
10.
by jwr 13d ago
If there is any company that can "ban your account" which would result in big problems for you (because, for example, they host your E-mail which means you've just lost access to password recovery for every account you have)
11.
by jwr 14d ago
I think before we get to self-referentiality, we have a bigger problem to solve, and that's one that I did not expect: namely, at which point do we stop saying that a machine pretends to be conscious and start saying that it is consc
12.
by jwr 15d ago
For what it's worth, I've spent some time with Claude to develop a local runner for `llama.cpp`. I run Qwen3.6-35B-A3B-MTP (fast!) and Qwen3.8-27B (20 tokens/s). This was definitely worth the effort. Benchmarking and testing
13.
by jwr 17d ago
I got so tired of every hardware device requiring cloud services, accounts, logins, privacy policies, etc. that I started building my own. I think the tipping point was a phone app to control a Panasonic air conditioner. It could just commu
14.
by jwr 19d ago
I spoke English and Polish to it. It produced Russian or Ukrainian as output.
15.
by jwr 20d ago
From what I understand, they also have no certifications (CE/FCC/VCCI and similar) and cannot really be sold as consumer products, only as kits.
16.
by jwr 20d ago
I used MacWhisper for a long time and then recently switched to the free TypeWhisper. Both work very well on a Mac.
17.
by jwr 20d ago
It's interesting that the results can be so different depending on the person, the use case and even the microphone used. I use dictation a lot, so I try to stay up to date with the latest models as much as I can. So far, for my needs,
18.
by jwr 20d ago
I use cLaude cOde in my eMacs (through ghostel, which is unbelievably great!) to generate code :-) I then use mAgit to review it.
19.
by jwr 22d ago
The problem is when you don't want to send your data to any provider. Private documents, health information, all that is something you don't want to send anywhere, even if the provider makes a pinky promise that they won't us
20.
by jwr 22d ago
Shipping with their computers -- maybe not, but they did make FoundationDB open source (again) and are developing and maintaining it.
21.
by jwr 23d ago
Hmm. What would be really helpful for me would be an enhancement of the TypeWhisper app for dictation that would be able to get the context of what I am dictating into and send it along with the prompt to an LLM. Probably much simpler and m
22.
by jwr 24d ago
I think synchronized operations could really disrupt the grid — think thousands of inverters switching backfeeding to the grid on/off, causing waves and disruption. Or thousands of inverters having their grid sync disabled, simultaneou
23.
by jwr 26d ago
Unfortunately, paying money to an economy of a country which started and still wages an unprovoked war against another country, killing civilians indiscriminately, is a choice. The only reason you can "focus on building great software&
24.
by jwr 26d ago
Ghostel is fantastic for this. It is very fast, fully compatible, and you can quickly enter a copy-paste mode with C-c C-t. With ace-jump this makes for very efficient copying. Since I started using Ghostel, my Claude Code and Codex sessi
25.
by jwr 27d ago
I have been using Emacs for 32 years now. It had its ups and downs: there was a period of very slow development, which led to the XEmacs fork, which I used for many years. These days I see a renaissance of sorts. First, the development prog
26.
by jwr 27d ago
This is… not obvious. It's only easy if you want the simplistic approach of sorting lexicographically or numerically. But in more complex tables, you will often encounter cases where the data representation for sorting is different fro
27.
by jwr 27d ago
Hmm, perhaps I should switch to an MLX version… problem is, it took quite a bit of work to get llama-server (with llama.cpp) to serve my model(s) and allow requests in non-thinking (default) and thinking modes. But 30-40 tokens/s would
28.
by jwr 27d ago
> Eventually I used Gemini 3.5 Flash for pairwise evaluation But, but… wouldn't that be… (gasp) DISTILLATION? Fun project!
29.
by jwr 27d ago
These are very good! I'm hoping for speed improvements because the only problem running the 27B model on my Macbook pro (M4 Max) is the speed: 20 tokens per second. I benchmarked and MTP actually makes things slower, so I disabled MTP
30.
by jwr 28d ago
I would suggest careful benchmarking. I actually tested and benchmarked, and the new Qwen3.8-27B model is actually slower with MTP on my M4 Max. MTP only gains anything when generating long code sequences, which is very unlikely as the mo

More ›