Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
JSR_FDED
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
5 ms
·
1.
▲
by
JSR_FDED
5d ago
Turtles all the way down?
2.
▲
by
JSR_FDED
8d ago
Clear-eyed description of the reality of corporate life. Designers feel this pain, as does anyone (like sw devs) who has a sense of “I could make this better” but isn’t given the support to do it.
3.
▲
by
JSR_FDED
8d ago
You don’t need anyone’s permission. You have only so much attention, why spend it wading through slop?
4.
▲
by
JSR_FDED
8d ago
Except that wasting the reader’s time became a lot easier with LLMs.
5.
▲
by
JSR_FDED
10d ago
Clutch?
6.
▲
by
JSR_FDED
13d ago
Repetitive Strain Injury inverts this statement
7.
▲
Is This the Future of America?
(theverge.com)
6 points
by
JSR_FDED
14d ago
|
2 comments
8.
▲
by
JSR_FDED
20d ago
Singapore too
9.
▲
by
JSR_FDED
21d ago
This could get messy
10.
▲
by
JSR_FDED
21d ago
I’m not religious but I’m fully on board with this pope’s take on human dignity, and worth. This is a moral, not a political and certainly not a tech issue. It’s good to have someone with a large megaphone weigh in on this who can’t immedia
11.
▲
by
JSR_FDED
22d ago
I think the blind test is valuable. To what degree is our impression of a model affected by who created it?
12.
▲
by
JSR_FDED
22d ago
There’s more than 2 LLMs in the world. Now whose post is stupid?
13.
▲
by
JSR_FDED
23d ago
In the Netherlands this has been my experience many times. By the time they finally come to take your order (which is always "what would you like to drink?", requiring me to say "actually I'm ready to order the food as w
14.
▲
by
JSR_FDED
23d ago
You can learn from the variety of techniques they used to come to this conclusion.
15.
▲
by
JSR_FDED
23d ago
Cool that they did this on top of Qwen3.6-35B-A3B. If they have their own collection of valuable data this is the only way to make sure it doesn’t end up in general purpose models. That’s probably enough justification for the $40m spend - c
16.
▲
by
JSR_FDED
23d ago
Interesting analysis. I wonder if the Chinese skills are undercounted somewhat- there could be more repositories the researchers didn’t have access to.
17.
▲
by
JSR_FDED
24d ago
You can use the ‘defaults’ CLI to manage settings files for the OS and applications. For example: https://macos-defaults.com/
18.
▲
by
JSR_FDED
24d ago
Listened to DHH on The Primeagen’s The Standup podcast - talking about how Macs are complete trash and can’t be used “for agents”. Like a dumb marketing droid parroting some inane talking points his VP has come up with to sound “AI relevant
19.
▲
by
JSR_FDED
24d ago
DHH is like Elon - at some point the personal brand tarnishes the project.
20.
▲
by
JSR_FDED
24d ago
But with systemd you can start deportation procedures while waiting for the rest to resolve.
21.
▲
by
JSR_FDED
24d ago
Because Simon usually posts interesting comments. Do I need a different reason?
22.
▲
by
JSR_FDED
24d ago
I’ve read the link, but I don’t get how it’s relevant to the OP’s point that Anthropic public communications don’t sound like Claude. Can you enlighten me?
23.
▲
by
JSR_FDED
24d ago
Other than the specific phrases like load-bearing etc, I find the biggest tell of all just to be repetition. Every damn Claude article does the “tell ‘em what you’re going to tell ‘em, tell ‘em, tell ‘em what you told ‘em” routine.
24.
▲
by
JSR_FDED
24d ago
I appreciate the inventiveness of this idea. But I don’t find myself thinking I must have this.
25.
▲
by
JSR_FDED
24d ago
Don’t forget a16z’s breathless boosterism of NFTs and Crypto. They really are the gift that keeps on giving.
26.
▲
by
JSR_FDED
24d ago
Honestly? That’s the kicker.
27.
▲
by
JSR_FDED
24d ago
No, it’s _Goodbye_ and thanks for the memories! Everyone knows that.
28.
▲
by
JSR_FDED
24d ago
But then there’s inevitably one great comment in there too, that suddenly makes things click or shifts your perspective…
29.
▲
by
JSR_FDED
24d ago
Agree! LLMs are interesting and useful, but this is creative !
30.
▲
by
JSR_FDED
24d ago
Would be great to also include some of the idiosyncrasies in how HN comments/discussions often develop. The way the first comment is often about something very tangential to the main topic of the post, and then proceeds to attract a la
More ›