Cakez0r

5 ms·

1.
by Cakez0r 3d ago
Fwiw I have a Samsung AC system and Claude was able to one-shot a custom web dashboard to control it.
2.
by Cakez0r 4d ago
I've been trying Grok 4.6 (High) alongside Claude Opus 5 (High). Grok is capable, but disappointing compared to Opus. My experience with Grok has also made me a little skeptical of benchmarks, because the two models benchmark similarly
3.
by Cakez0r 4d ago
Just like that! I will give it a try
4.
by Cakez0r 5d ago
It would be interesting to see a decentralised, residential collective that builds and publishes an index. There are surely enough interested people on HN alone that would be willing to run software at home to scrape a small slice of the in
5.
by Cakez0r 5d ago
If it's true just link to a first party example on the yandex site
6.
by Cakez0r 18d ago
> So how do you migrate 2.4 million lines of code? > Very carefully, and in multiple stages. That answer is so 3 months ago
7.
by Cakez0r 21d ago
They're both the same root cause. Command injection in notifications
8.
by Cakez0r 21d ago
The two links are the same same bug. Hardly "full of security holes"
9.
by Cakez0r 21d ago
"full of security holes" but the author could only name one (the two links in the opening paragraph are the same issue). Some people just hate DHH and can't separate the art from the artist
10.
by Cakez0r 26d ago
Omarchy does all of those things
11.
by Cakez0r 26d ago
The hype is real though. It's a great distro
12.
by Cakez0r 26d ago
It was (pretty much) just dot files and curated software, although I think that description downplays the value of having a perfectly tuned system out of the box - like saying a log cabin is just a bunch of sticks. The assembly matters and
13.
by Cakez0r 26d ago
What don't you like about Omarchy? I'm not young and I used Fedora before Omarchy. I think it's excellent
14.
by Cakez0r 1mo ago
This could be why claude code has recently started to write reams of inane comments alongside the code it generates.
15.
by Cakez0r 1mo ago
I think you're right that it's actually to do with attribution (i.e. They're not just watermarking your output as claude generated, but watermarking it as claude generated _by claude user id 73684_). I think it's more to
16.
by Cakez0r 1mo ago
Yes It falls from the sky Septic tank
17.
by Cakez0r 2mo ago
The internet is like a militarised zone now. Joining a new website feels like going through airport security and being on the modern web feels like being stuck in an economy seat on a budget airline.
18.
by Cakez0r 2mo ago
If you don't live in a first world country then you will also find yourself on the "too bad, don't care" list.
19.
by Cakez0r 2mo ago
In their defence, don't shoot the messenger. Just because they published it doesn't mean that others haven't already discovered it. Better to know its possible than be completely ignorant.
20.
by Cakez0r 3mo ago
I'm seeing that the headline chart of the article is inexplicably deceptive. Why not use coloured data points and a legend? If you're going to have logos with lines pointing to the data points, why make the data points and lines f
21.
by Cakez0r 3mo ago
LLMs spotted https://hackernewstrends.com/?q=Genuinely
22.
by Cakez0r 3mo ago
All the chatbots refused to confirm your political bias. Seems like a win to me
23.
by Cakez0r 3mo ago
The overall chart on their page seems politically biased, or at the very least a chart crime. The logos have faint grey lines pointing to their actual data points, conspicuously making Grok look very far right. You wouldn't know it fro
24.
by Cakez0r 3mo ago
> The next time someone talks about restricting immigration, show them this: Nearly half of America’s billion-dollar startups were founded by people born outside the United States. I don't think the people that talk about restrictin
25.
by Cakez0r 3mo ago
Mimo 2.5 pro is the best intelligence / dollar
26.
by Cakez0r 3mo ago
I'm not sure about those parameter sizing claims. Regardless of parameter size, benchmarked intelligence of Chinese and Western frontier models is comparable, so who cares how many parameters it takes to get there. Mimo is also widely
27.
by Cakez0r 3mo ago
0/10 succesful attempts for mimo v2.5 pro (high) using opencode. It was not able to think bigger than exploiting vectors outside of the API. However, I felt the prompt was implying that only authenticated API requests are fair game, so
28.
by Cakez0r 3mo ago
Here are 3 benchmarks showing the comparable scores I was talking about https://openrouter.ai/rankings https://arena.ai/leaderboard/text/coding https://artificialanalysis.ai/
29.
by Cakez0r 4mo ago
It is totally slept on. In my experience it is cheap, fast and capable (not just capable with caveats, but just as capable as western flagships). My only gripe with it is that sometimes the API seems to timeout which tanks the overall speed
30.
by Cakez0r 4mo ago
It would be interesting to see full results for Kimi K2.6 and Mimo v2.5 pro. These two models benchmark comparably to other flagship models. Having these complete results would give a clearer picture of the AI frontier. EDIT: I have a mimo

More ›