croemer

48 ms·

481.
by croemer 1y ago
This claim is talking out of proportion, LLMs do push back: > Input Risk. An LLM does not challenge a prompt which is leading or whose assumptions are flawed or context is incomplete. Example: An engineer prompts, "Provide a thread-
482.
by croemer 1y ago
Is there any evidence that it actually uses Markov chains? I don't see any mention. The title here might as well be made up to troll HN?
483.
by croemer 1y ago
> You will be Ctrl + FrontMouseScrollWheel to read your 3rd monitor too eventually.. I'm failing to parse this
484.
by croemer 1y ago
The mobile website is different than the desktop version, that's not very common anymore. There's no GitHub logo in the mobile version.
485.
by croemer 1y ago
Parallelism isn't speed. Zstandard decompression is insanely fast, no need for occupying multiple cores.
486.
by croemer 1y ago
Can you share this chat via Claude's sharing feature? I'm skeptical and can't reproduce: https://claude.ai/share/c1454ae7-a2ee-4067-9b54-484340d36b4e
487.
by croemer 1y ago
Indeed, as this comparison with gzip shows: https://github.com/ross39/new_bloom_filter_repo/blob/main/co...
488.
by croemer 1y ago
Why would it be faster? You have to do all that hashing and lookup of witness data etc. Also, if you want fast, good compression, use zstandard not gzip.
489.
by croemer 1y ago
The website looks like 2013 and much of the content is as well. There's a GitHub repo that I couldn't find from the website: https://github.com/red/red
490.
Vibe coding: your roadmap to becoming an AI developer(github.blog)
3 points by croemer 1y ago|0 comments
491.
by croemer 1y ago
Title needs a dash after Google, otherwise it reads weirdly
492.
by croemer 1y ago
Discussion on Meta SO: https://meta.stackoverflow.com/questions/433864/do-you-agree...
493.
by croemer 1y ago
Questions asked isn't the metric to declare it dead. What matters for financial viability is traffic and deals with search engines/indexes (used by LLMs) For community viability: people will keep using it where LLMs fail. For new
494.
by croemer 1y ago
More typos (LLMs are really good at finding these): "Press the "Run Simulation" button to start traininng the RBM." ("traininng" -> "training") "...we want to derivce the contrastive divergenc
495.
by croemer 1y ago
The blog post is tagged with "improvement" - ironic for more restrictive rate limits. Also, neither the new nor the old rate limits are mentioned.
496.
by croemer 1y ago
Indeed, article appears to be old encyclopedia style, no citations, oversimplified.
497.
by croemer 1y ago
Right, I got the direction wrong.
498.
by croemer 1y ago
Involved only insofar as shipping it as setuid-root counts. Distros that configure it like this are vulnerable, others aren't. Very thin involvement I'd say. Distros patch if upstream is too slow.
499.
by croemer 1y ago
What the article doesn't emphasize enough: Pinning of the contact line is crucial (e.g. due to surface roughness), otherwise the ring would not be as pronounced. Due to higher curvature, evaporation is faster at the edges, causing the
500.
by croemer 1y ago
What do you mean by "fe"? "For example "?
501.
by croemer 1y ago
You've made me curious about the tapeworm theme!
502.
by croemer 1y ago
> We have also updated the model card with the new version of 2.5 Pro No you haven't? At least not at 6am UTC on May 7. The PDF still mentions (03-25) as date of the model. What version do I get on gemini.google.com when I select &q
503.
by croemer 1y ago
But the reverse isn't true, right? It doesn't prevent the board from maximizing financial returns even when doing so would harm the "public".
504.
by croemer 1y ago
Nope, people don't need to know that something is done to appreciate the outcome. You might not know that modern MacBooks use ARM processors, but you might still appreciate that they have a long battery life.
505.
by croemer 1y ago
The MDM admin can do whatever the user can do (or more), sure. So yes the MDM admin can potentially read/hear/see stuff, but everyone knows that. That's not a vulnerability, that's by design. The compromise is only wrt t
506.
by croemer 1y ago
https://archive.is/6J8mf
507.
by croemer 1y ago
See this for a non-flagged discussion on the topic: https://news.ycombinator.com/item?id=43875476
508.
by croemer 1y ago
The fact that archive link works should make this eligible for unflagging. From tomhow (mod) > If it's a good article (contains significant new information and can be a topic of curious conversation) and a paywall workaround works f
509.
by croemer 1y ago
Why do you say "everything said on those phones" - did you mean "on this app"? If the backend of an app was compromised, that wouldn't mean the phone itself was rooted?
510.
by croemer 1y ago
> But alas. We won’t get any of that. We can still comment and read, it's just not shown by default. I found it by submitting it myself just to find it flagged.

More ›