jdiff

14 ms·

181.
by jdiff 3mo ago
Not to mention the very first example. Before: > “I want to know why my React app’s state is not updating when I click a button.” After: > “React 18. useState. Button click handler sets state but component does not re-render. No error
182.
by jdiff 3mo ago
There's always the next great kernel level font or scrollbar exploit.
183.
by jdiff 3mo ago
Yes. In this case, during fine tuning. Other blurbs are also baked in during fine tuning that are perfectly reproducible from the Nex model. The details inside the linked issue are quite accessible.
184.
by jdiff 3mo ago
This is contrary to the mayor's words on Twitter. > An open AI model trained in Rio with public funding over the last year by @Prefeitura_Rio surpassing all other models. https://x.com/CavaliereRio/status/2
185.
by jdiff 3mo ago
Without the system prompt, asking its name results in it responding with the name of the model they're ripping from. That would certainly draw your eyes to the right places.
186.
by jdiff 3mo ago
That shouldn't be used to judge other models - it's never been true for Grok.
187.
by jdiff 3mo ago
Smear frames are not something often applicable in this kind of animation. Smear frames are just about specific to frame-by-frame animation. No smear frames are demonstrated in this article.
188.
by jdiff 3mo ago
You're thinking of smear frames. Squash and stretch are animation techniques that are perfectly coherent. Smear frames as well contribute to an overall coherent animation. They're a counterpoint to the general idea put forward in
189.
by jdiff 3mo ago
I have a coworker who, when he needs to operate some software that is unfamiliar to him, snaps a photo of it and has Gemini AI read each label and description. If there is a checklist or form that needs to be filled, Gemini reads each quest
190.
by jdiff 3mo ago
It's been happening for some time now. I saw this pop up over 6 months ago. After searching I can find some dating back at least a year.
191.
by jdiff 3mo ago
LLMs don't make a good A-frame, nor would I classify them as wood-like. People propose LLMs as solutions as if they're wooden when they're teetering contraptions of metal rods, aluminum extrusions, rubber bands, and duct tape
192.
by jdiff 3mo ago
From elsewhere in the thread, some hard numbers on the topic. https://formulae.brew.sh/analytics/homebrew-os-arch-ci/30d/ Intel homebrew is larger than Linuxbrew, yet I think it'd be shocking if they dro
193.
by jdiff 3mo ago
Had a little luck with having it do an impression of the Star Trek computer, although at the cost of having it try to insert star-trek themed hallucinations like warp engine status.
194.
by jdiff 3mo ago
We already have that in the form of separate reasoning/thinking and speaking streams. Even with that it's awfully hard to get LLMs to keep it consistently concise. As soon as that context window starts growing it falls right back
195.
by jdiff 3mo ago
OpenCode seemed perfectly workable as a programming assistant. As personal assistants, they all fall short. It's too difficult to really shape their output. I was briefly impressed with OpenClaw a few times, but ultimately was turned o
196.
by jdiff 3mo ago
There's always an excuse, or perhaps someone who actually went through it might be aware of the process and limitations better than your surface level assumptions?
197.
by jdiff 3mo ago
Appreciate the concept, seems deeply useful if a bit underbaked at present. Active STT allows a "No STT loaded" option that mentions it requires a multimodal LLM like Gemma 4. Except even when I use Gemma 4 features, Ctrl+S to dic
198.
by jdiff 4mo ago
Atrophy of your basic skills can still be a problem. Like someone relying on a dozen specialized triangle angle calculators found on Google rather than understanding what SOHCAHTOA is.
199.
by jdiff 4mo ago
Not the same person, but in the event of an AI collapse I think those that relied on it will be at a disadvantage. The rapid deskilling that happens with AI usage is becoming more documented.
200.
by jdiff 4mo ago
It's nearly a line-by-line rephrasing of another story that's linked at the top. In the second half it takes less creative liberties and sticks closer and closer to the original meat-based text.
201.
by jdiff 4mo ago
Most package managers with postinstall scripts are also heavily curated and have reputation systems. As you say, they run as root, so the high trust requirement is definitely warranted. Anyone can upload an npm package.
202.
by jdiff 4mo ago
The operator highlights "Don't stand down" and "Champion free speech" but the thing that grabs my eyes is right at the top, the typo and the heady ego of "programming God!" Everything in the context will g
203.
by jdiff 4mo ago
This makes no sense given development is not driven by any one entity that might work privately and start publishing later. All development on these projects is done in the open by a variety of entities who have no mutual interest in collud
204.
by jdiff 4mo ago
So obviously none of this is realistic, even the foundations of it don't align with reality, so what point are you trying to make with this comment? I'm afraid it's going over my head.
205.
by jdiff 4mo ago
A merge to main itself is pretty substantial, especially a week after saying, "[This] code that does not work. We haven’t committed to rewriting. There’s a very high chance all this code gets thrown out completely."
206.
by jdiff 4mo ago
I think any instruction tuned model is going to "know" it's an LLM.
207.
by jdiff 4mo ago
I've never had much contact with Mercurial myself, so are there any features from Mercurial that JJ doesn't already incorporate? Or any differences you find interesting?
208.
by jdiff 4mo ago
The same crowd who always vote against their own interests. To paraphrase, "give them somebody to look down on, and they'll empty their pockets for you."
209.
by jdiff 4mo ago
Have you used Jujutsu before? It's git-backed and it sounds like it incorporates a lot of these niceties from Mercurial. I find it an awful lot more intuitive than Git to use and the stable identifiers are absolutely lovely to have.
210.
by jdiff 4mo ago
Memory bandwidth is completely different on any browser from measured results on my M2 Pro machine. Weirdly, the estimated performance levels and even exact product name differ between Chrome and Firefox. Firefox calls it an M2 Pro and over

More ›