skerit

9 ms·

151.
by skerit 1y ago
That's why people called the second version of Sonnet v3.5 simply v3.6, and Anthropic acknowledged that by naming the next version v3.7
152.
by skerit 1y ago
Strictly 2 spaces for indentation? What's human about that.
153.
by skerit 1y ago
> while the Americas (generally) assign house numbers based on the distance to the beginning of the block What? That sounds great! So if you're at house number 247 could you deduce, in meters, how far away house 1483 is?
154.
by skerit 1y ago
People can make utter trash on Bake Off, get honest critiques, and yet the vibe is _still_ good. It's really great. Now a bit off-topic: Bake Off has completely put me off baking myself. Nearly everything always goes wrong. And as this
155.
by skerit 1y ago
I don't get it. Tools are a way to let LLMs do something via what is essentially an API. Is it limited? Yes, it is. By design. Sure in some cases it might be overkill and letting the assistant write & execute plain code might be be
156.
by skerit 1y ago
I read the introductory post but I still don't quite understand
157.
by skerit 1y ago
Indeed, all of the above. It's impossible to just say "Here is the original file, here is the new file, port everything over" and think that'll be enough. Maybe making a plan for each file and specifically tell it to por
158.
by skerit 1y ago
This is indeed super annoying. I always have to add something like "Don't do anything just yet, but could it be ..."
159.
by skerit 1y ago
I remember downloading the last 2 episodes of Star Trek Voyager of of an FTP server through Windows XP's Explorer. 2 WMV files of 50 megabytes each. Took all night. If a single issue had happened I would have had to start the download
160.
by skerit 1y ago
Incredible that we're getting something like this before a plain good old screensaver
161.
by skerit 1y ago
I like using Claude-Code, it can be a real timesaver in certain cases. But it's far from perfect. Really difficult things/big projects are nearly impossible. Even if you break it down into hundred small tasks. I've tried to m
162.
by skerit 1y ago
I don't understand the example image. The description says: > The yellow circles highlight areas of similarity between the original photo and the AI-generated photo What about the person in the middle, doing the throwing? It's
163.
by skerit 1y ago
A lot of people use LLMs for fiction & role playing. Do you know of a place where some of these interactions are shared? The only ones I've found so far are, well, over-the-top sexual in nature. And I'm just kind of interested
164.
by skerit 1y ago
In my experience, Gemini is pretty good in multishotting. So just give it a system prompt, some example user/assistant pairs, and it can produce great results! And this is its biggest weakness for coding. As soon as it makes a single m
165.
by skerit 1y ago
I agree. They have some good reasons on why AI-generated code should not be used in this project, but the page just devolves into a "All AI is bad" and the constant use of "Slop generator" just makes me think of all the
166.
by skerit 1y ago
I don't care how good it is, I'm not spending money on any of Elon Musk's products.
167.
by skerit 1y ago
40,3399 BEF They really hammered this in. It was even a question on my exams.
168.
by skerit 1y ago
> Larry has lived at 10 Downing Street during the premierships of six prime ministers Six! The troublesome times this cat has witnessed from close by...
169.
by skerit 1y ago
What is this thread? Sticking a knife through your skin to the bone kills pain and is addictive?
170.
by skerit 1y ago
I tried it too, it was so bad. I got the same "revert" behaviour after only 15 minutes.
171.
by skerit 1y ago
Google announced a Gemini CLI tool a few hours ago! And then ... they unpublished everything shortly after? It's weird: https://news.ycombinator.com/item?id=44373754
172.
by skerit 1y ago
Another bonus for Claude Code: with the max plans, you don't have to worry about per-token costs. Sometimes a certain attempt does not pan out, but I don't have to feel bad because it didn't cost me hundreds of dollars for it
173.
by skerit 1y ago
LLMs aren't necessarily taught the characters their tokens represent. It's kind of the same how some humans are able to speak a language, but not write it. We are basically "transcribing" what LLMs are saying into text.
174.
by skerit 1y ago
I'm sorry, I really like it. When used in titles & subtitles, I find it makes it a lot more pleasant to read for me.
175.
by skerit 1y ago
I am kind of happy that people are looking back at iOS 7's style & flat design in a whole and can just say: "I hated it then, I hate it now". Because I surely do. Back then if you hated it, you were just being stubborn, y
176.
by skerit 1y ago
> I don’t think I’ve encountered a case where I’ve just let the LLM churn for more than a few minutes and gotten a good result. Is this with something like Aider or CLine? I've been using Claude-Code (with a Max plan, so I don'
177.
by skerit 1y ago
Could be, but getting a Gigsky e-sim is a lot cheaper than the internet on-board. And I was under the impression that in the end, that system is using the same internet uplink as the wifi (since I believe it's the ships themselves that
178.
by skerit 1y ago
Sonnet or Opus? Well, I guess they both still can do that. But I'm just keeping on asking it to review all its code. To make sure it works. Eventually, it'll catch its errors. Now this isn't a viable way of working if you&#x2
179.
by skerit 1y ago
Sometimes it's great, sometimes it's not. Depends on the tools you're using too, I guess. Like when using Roo-Code, Gemini 2.5 Pro still gets confused by the wonky diff format Roo-Code wants it to use. It'll keep messi
180.
by skerit 1y ago
I've been using Claude-Code for a few weeks now, and honestly, calling this just "ergonomic" tasks feels like a huge understatement. This thing is straight up writing code for me - real, functional code that actually works. I

More ›