rpdillon

9 ms·

211.
by rpdillon 2mo ago
This is a copy-pasted comment from elsewhere in the thread, which I've already responded to. A bit much calling others "incredibly lazy".
212.
by rpdillon 2mo ago
Not at all. The post being responded to completely failed to acknowledge any value brought by data centers to the general populous. That's disingenuous. To force the point: run me through the scenario where we turn off all the data cen
213.
by rpdillon 2mo ago
Why would they? They're not the ones upset that they're cutting up books. They fought a court case to allow them to do it a different way that didn't require cutting up any books. They were shut down and had to pay billions,
214.
by rpdillon 2mo ago
You've jumped from "I don't understand why anybody would do this" to "I don't want it". People are different. People have different use cases. People live in different areas.
215.
by rpdillon 2mo ago
This just isn't true. You're assuming the conclusion and saying the routes will have equal amounts of traffic and that one will be way longer and that it's unlikely you're saving time. Except people that have experience
216.
by rpdillon 2mo ago
Ah, thanks, I'd missed that nuance in the other reply!
217.
by rpdillon 2mo ago
Very much appreciate you driving this home. I had penned a lengthy response about relative timescales, but you summed it up nicely.
218.
by rpdillon 2mo ago
A billion tokens a day is 11,000 tokens a second sustained. How many tokens per second are you getting off of GPT 5.6 Sol per project?
219.
by rpdillon 2mo ago
Right, but that's because of copyright law.
220.
by rpdillon 2mo ago
Yes, we've inadvertently set up a system of incentives that were designed to preserve information and monetize it. And we've instead set up a set of incentives to make it scarce and destroy it.
221.
by rpdillon 2mo ago
> I would say the burden of proof is on the companies destroying human cultural heritage en masse, not the handful of journalists calling for attention to the matter. You're literally assuming the conclusion. The exact topic under c
222.
by rpdillon 2mo ago
The titles mentioned there are not rage-inducing: "The Insider’s Guide to Metro Denver from 1995" and "How to Use Corel WordPerfect 1991". 404 writes and article about book destruction, and then explicitly neglects to me
223.
by rpdillon 2mo ago
Seems like getting a copy and scanning it and releasing it as part of a digital library would be a good way to preserve the knowledge.
224.
by rpdillon 2mo ago
Can we copy the books? It seems like in an effort to monetize writing, we've created a system that incentivizes destroying it.
225.
by rpdillon 2mo ago
All of the images in the article that are not of the vandalism signs are, "representative images" that they pulled from Getty Images. They have zero content because they have nothing to do with the article.
226.
by rpdillon 2mo ago
Yeah, I think they are strays added by GitHub Actions, they don't line up with the actual version that's tracked inside the source code, which indicates v0.20.1: https://github.com/ggml-org/llama.cpp/comm
227.
by rpdillon 2mo ago
Yeah, it's an interesting test because he gave Kimi credit for running inside the harness he liked, which I presume is Claude Code. But he also said it was noted for its harness sensitivity, and then said it didn't do as well as O
228.
by rpdillon 2mo ago
In a situation where you are an editor and focusing on the clarity and the brevity of the writing, I think this is a fantastic suggestion. Hacker News has long ago decided that making comments on the quality of the writing in the comments i
229.
by rpdillon 2mo ago
And those examples were not cases where companies went back to manual work after using AI. They were examples of companies regretting laying off their experts. The point GP is making, which I understand seems dismissive, is that AI is a too
230.
by rpdillon 2mo ago
This is correct. It's a giant red flag in GP that they're worried about their developers making a mess with AI. That's an issue of the developers' judgment not being any good, not a problem with AI.
231.
by rpdillon 2mo ago
Ford did not . They hired back veterans to better train the AI. > COO Kumar Galhotra said Ford had been over-relying on automated quality systems without getting results, per Bloomberg. The returning engineers rebuilt the data pipelines
232.
by rpdillon 2mo ago
Fairly brittle approach: https://codeberg.org/polyphony/repo-slopscore Their methodology section says they scan for authors and commit messages indicating AI use. This wouldn't work on any of my repos, for example
233.
by rpdillon 2mo ago
If the concern is that you're touching people with a piece of metal, it matters whether it's electrified. We've essentially invented electricity from the 20-year gap you're discussing, and you're saying that we shou
234.
by rpdillon 2mo ago
This is a dramatic oversimplification to the degree that I do not think the conclusions can be drawn from the assertions. I responded to you above where you mentioned that AI cannot do nuance with some information from the judge in this par
235.
by rpdillon 2mo ago
The judge in his rulings talks significantly about the patterns of use here, and pins it squarely on the AI being used by low skilled individuals. In particular, he points out that they typically ask AI to support their position and make an
236.
by rpdillon 2mo ago
The proof is the judge said he doesn't run his courtroom that way.
237.
by rpdillon 2mo ago
No, they don't. That was the author of the article using something the judge said in reference to AI court systems elsewhere and applying it to a case where it did not apply because no AI was in use. There's a long post near the t
238.
by rpdillon 2mo ago
Boox is fantastic. I have the Palma and it's been very enjoyable, since it's pocketable and has excellent power management, so despite being android, the battery can last 1-2 weeks.
239.
by rpdillon 2mo ago
Can't edit my post anymore, but here's the omp blog from February talking about improved tool calling rates across 15 models, with only the harness being tweaked to get the improvements. https://stencil.so/blog&#x2
240.
by rpdillon 2mo ago
Tool calling success rate, in the case of omp.

More ›