Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
sebastiennight
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
31.
▲
by
sebastiennight
1mo ago
This point makes sense to me. I agree with "learned" and "discovered".
32.
▲
by
sebastiennight
1mo ago
After a bit of thinking, I agree that "learned" is the most appropriate. I stand corrected.
33.
▲
by
sebastiennight
1mo ago
You wrote your house's security code on a Post-It note. Somebody came in as a houseparty guest and located the note, and discretely copied it onto their hand to get access to the house during your next vacation. Which verb would descri
34.
▲
by
sebastiennight
1mo ago
> Terseness doesn't offend the statistical model, but giving platitudes may color its output My understanding of the state of current research is that the impact of tone/politeness on performance is highly model-dependent, and
35.
▲
by
sebastiennight
1mo ago
You're forgetting some detail in your example, which is that 1. the existing STORES also sell almost all the CARS, and the GPS in these cars is hardcoded to only drive to one of their STORES 2. The government is considering a law that
36.
▲
by
sebastiennight
1mo ago
Nice execution, and though I was assuming the website would also be claude-generated, but it avoids a lot of the "AI look and feel". How did you prompt the writing and design? A nitpick just because this sentence had me pause for
37.
▲
by
sebastiennight
1mo ago
I would have said the opposite: I'd assume the link to be to a blog post, or Wikipedia page, that articulates the idea in a clearer more succint way than the person might be able of. This is, of course, before people started sending In
38.
▲
by
sebastiennight
1mo ago
Counterpoint: if you're uncool yourself, you might be uncool enough to call out uncoolness, from a "game recognizes game" point of view
39.
▲
by
sebastiennight
1mo ago
Am I the only one who's had Claude almost systematically remove human-written comments? It might be touching one line of actual code in a file, and take advantage of it to remove 20+ lines of actual useful comments. Everybody is talk
40.
▲
by
sebastiennight
1mo ago
Yes, I meant, find the closest word by edit distance. I'm unsure where vectorization would factor in here?
41.
▲
by
sebastiennight
1mo ago
> it says the same thing Reference needed? I think it remains to be proven whether those detectors can be considered deterministic.
42.
▲
by
sebastiennight
1mo ago
Saying "half" relies on a lot of assumptions, as either side of that count can be made arbitrarily larger or smaller based on how many items you want to subdivide it into.
43.
▲
by
sebastiennight
1mo ago
This phenomenon is called the Euphemism Treadmill: https://en.wikipedia.org/wiki/Euphemism_treadmill
44.
▲
by
sebastiennight
1mo ago
Historically, no. The simplest way to build autocorrect is to have a dictionary, look it up for each word, and suggest the "closest" dictionary word if a typed word is absent from the dictionary. It has existed way before transfor
45.
▲
by
sebastiennight
1mo ago
My understanding is that they gave up the USSR's weapons (and handed them to Russia), rather than them having built, owned and dismantled the weapons outright through their own nuclear program. They could have spent substantial effor
46.
▲
by
sebastiennight
1mo ago
A public service reminder to anyone setting up their domains that there are many alternatives to Cloudflare, like EG Bunny.net (I am not affiliated, just a happy customer) and that by using one of these other DNS providers, you - play your
47.
▲
by
sebastiennight
1mo ago
"can't" is incorrect. As your name server, they can decide to turn the reverse proxy/caching at any point in the future, write an obscure blog post about how this is the best solution for all (a post that gets 2 comments
48.
▲
by
sebastiennight
1mo ago
I knew about it, but your post is the first I've heard that SA no longer had those (I only clicked the link because you used the past tense!). I actually thought that no country had ever withdrawn their nuclear program. This might be
49.
▲
by
sebastiennight
1mo ago
I came here looking for that information, as I was incredibly surprised to learn of nuclear energy existing in the Caribbean. It would be a dream to no longer have to rely on fossil fuels, but my native island is at the peak possible cross-
50.
▲
by
sebastiennight
1mo ago
I think GP is using a different level of abstraction from yours in their metaphor. You are saying: Pre-Training == Everything you store in your memory throughout your life. Model weights == The lessons you learned Context == whatever you&#x
51.
▲
by
sebastiennight
1mo ago
It also seeps into all documents and artefacts it creates. Claude will include actual comments ("// ...") into Excel sheets, and include the thinking that led to the output, instead of just focusing on the final result.
52.
▲
by
sebastiennight
1mo ago
With proprietary labs lowering their prices and Deepseek raising theirs over time, wouldn't it possible to extrapolate a graph to look at where the terminal frontier-model million-token-cost asymptotes to?
53.
▲
by
sebastiennight
1mo ago
> Because people are just using AI for search now (this is the main reason I use it) that there will be no point in making these spamming sites anymore Or: because people are just using AI for search now, the only websites left will be
54.
▲
by
sebastiennight
1mo ago
> your life is still going on but its going on in the forward direction, not backwards in time to overlap an earlier HN state Wouldn't the proper tense be "I won't have lived long enough to see the entire lifetime of HN&qu
55.
▲
by
sebastiennight
1mo ago
Good point. And it would revolutionize our current understanding of physics to find out that p = p̄
56.
▲
by
sebastiennight
1mo ago
I didn't say "profitable viable software that turns into a unicorn"... but according to the OP article, it looks like they are realizing that until they figure out how to make these software demos successful, they'll str
57.
▲
by
sebastiennight
1mo ago
Counterpoint: Our engineers are currently trying to replicate "intelligence" and we think it's like the brain because today we think the brain is where "intelligence" reside, and today we strongly believe that
58.
▲
by
sebastiennight
1mo ago
One of the reasons this analogy is unconvincing is that humans have compared themselves and their inner workings to "the current technology of the time" for millenia: - ~3rd century BCE : The invention of hydraulic engineering (eg
59.
▲
by
sebastiennight
1mo ago
In a "clean room" style implementation, as well. We've looked at the outputs, wrote a spec, and built from there. In that sense, I guess at least LLMs are not infringeing on the kind of patents that would get the owners of AI
60.
▲
by
sebastiennight
1mo ago
> It's front and back that it swaps. This sub-thread is becoming unfairly more interesting than the main conversation at this point.
More ›