skerit

7 ms·

61.
by skerit 5mo ago
> I worked on this project intermittently for 10 years, until recent developments in LLMs finally made it possible to complete this seemingly never-ending task. I've been working on my own MFC C++ decompilation project. It's in
62.
by skerit 5mo ago
For years, even leading up to starting the decommission of the power plants, Engie has been saying it's literally impossible to reverse the decision. And now that we're 2 years into the decommission, suddenly it is possible after
63.
by skerit 6mo ago
~~That just changes it to Opus 4, not Opus 4.7~~ My statusline showed _Opus 4_, but it did indeed accept this line. I did change it to `/model claude-opus-4-7[1m]`, because it would pick the non-1M context model instead.
64.
by skerit 6mo ago
I like this in theory. I just hope it doesn't require you to be be as literal as if talking to a genie. But if it'll actually stick to the hard rules in the CLAUDE.md files, and if I don't have to add "DON'T DO ANYT
65.
by skerit 6mo ago
I'll believe in this miracle model when I see it.
66.
by skerit 6mo ago
It comes over as so incredibly insane to me that people from the late 80s (people working with computers! Reporting on them!) would look at their current technology stack and basically go: "I have no idea whatsoever what else we can do
67.
by skerit 6mo ago
I'm sure it'll be better than Opus 4.6, but so much of this seems hype. Escaping its sandbox, having to do "brain scans" because it's "hiding its true intent", bla bla bla. If it manages to work on my java
68.
by skerit 6mo ago
I have the 20x plan and use it together with my husband. 4 days in to our weekly usage window and we're only at 54% (and we both use it the entire day) I have no idea how people are hitting the limits so fast.
69.
by skerit 6mo ago
I retired my 5700 XT a few years ago. Wasn't there some kind of hardware problem with it? It kept locking up my Linux kernel.
70.
by skerit 7mo ago
Neat. Is it a single under-trained token in GPT-5.2? Or is something else going on?
71.
by skerit 7mo ago
I've been running my own (admittedly naïve) experiments of new, wacky ideas for both LLMs (well, SLMs) and for Image-Super-Resolution models. I'm just trying different kinds of attention mechanisms, different configurations of the
72.
by skerit 7mo ago
Do we also have stills of all the episodes? Or only audio?
73.
by skerit 7mo ago
This is kind of what LoopLM is doing, no? https://arxiv.org/abs/2510.25741
74.
by skerit 7mo ago
Gemini might be great at benchmarks, it is terrible at actual agentic coding. So Anthropic seems like a more logical choice.
75.
by skerit 7mo ago
I hate the new way scrolling works in the Youtube & Netflix apps. So janky.
76.
by skerit 8mo ago
> They were rare, and special, and you'd have a few photos per YEAR to look back on. The feel of photos back then, was at least 100x stronger than now. [...] But once they became freely available that same amount of emotion is now s
77.
by skerit 8mo ago
> Saying it's "fully conscious" is silly, and anyone with this background should know better I'm surprised that anyone that truly knows how LLMs work would ever think they're sentient. I made a little presentatio
78.
by skerit 8mo ago
> Then a brick hits you in the face when it dawns on you that all of our tools are dumping crazy amounts of non-relevant context into stdout thereby polluting your context windows. I've found that letting the agent write its own opt
79.
by skerit 8mo ago
> I used Claude Code and Codex for the translation. This was human-directed, not autonomous code generation. I decided what to port, in what order, and what the Rust code should look like. It was hundreds of small prompts, steering the a
80.
by skerit 8mo ago
Lol, and this line: > Geminin 3.1 Pro can comprehend vast datasets Someone was in a hurry to get this out the door.
81.
by skerit 8mo ago
I'm glad someone else is finally saying this, I've been mentioning this left and right and sometimes I feel like I'm going crazy that not more people are noticing it. Gemini can go off the rails SUPER easily. It just devolves
82.
by skerit 8mo ago
Nun, V frr. V npghnyyl gbbx n ybbx ng gur cybg naq V checbfrshyyl fcbvyrq gur raqvat bs gur fubj sbe zr, naq V jvyy abg or jngpuvat vg. V gubhtug guvf jbhyq or fbzr shaal zbpxhzragnel, ohg gur qenzn natyr jurer vg pbzcyrgryl ehvaf fbzrbar&#
83.
by skerit 8mo ago
Agreed! I don't understand how so many people on here seem to think it is completely reasonable for Anthropic to act like this.
84.
by skerit 8mo ago
Which other CLI agents are that? Because I've found OpenCode to be A LOT better than Claude-Code.
85.
by skerit 8mo ago
It would be less of an issue if Claude-Code was actually the best coding client, and would actually somehow reduce the amount of tokens used. But it's not. I get more things done with less tokens via OpenCode. And in the end, I hit 100
86.
by skerit 8mo ago
> Avoid seeing episode counts if at all possible I don't understand. How would knowing how many episodes there are have an impact on the show? (And the number of episodes seems just... fine?)
87.
by skerit 8mo ago
Oh, so does "Counterpart" not have an ending?
88.
by skerit 8mo ago
I've been doing some decompilation with Ghidra. Unfortunately, it's of a C++ game, which Ghidra isn't really great at. And thus Claude gets a bit confused about it all too. But all in all: it does work, and I've been abl
89.
by skerit 8mo ago
When did they successfully close the loophole? I know they tried a few times, but even the last attempt from a week or two ago was circumvented rather easily.
90.
by skerit 8mo ago
Indeed, Opencode has it too. They've been improving it the past few weeks to look more like the one in Claude-Code. I disable it all the time though, I find it such a pain (in both Claude-Code and OpenCode)

More ›