HackerTrans
TopNewTrendsCommentsPastAskShowJobs

ngruhn

671 karmajoined 7년 전

Submissions

[untitled]

1 points·by ngruhn·2개월 전·0 comments

What happens at 60% unemployment rate?

gruhn.me
9 points·by ngruhn·5개월 전·5 comments

Show HN: Generate Passwords from Regex Constraints

gruhn.github.io
8 points·by ngruhn·7개월 전·6 comments

Check equivalence of any two regex

gruhn.github.io
2 points·by ngruhn·9개월 전·0 comments

comments

ngruhn
·3일 전·discuss
That's a beautifully simple characterization. I've never heard of of that. What keywords do I need to find more on this topic?
ngruhn
·9일 전·discuss
> “Safety” in the context of LLMs concerns not offending certain individuals.

It absolutely concerns not destroying the world.
ngruhn
·9일 전·discuss
Imagine a nuclear power company being extremely cautious with safety and everyone online is like "make more power already you pussies!"
ngruhn
·10일 전·discuss
> I think the models are being optimized for wealth extraction from users and companies, instead of solving problems.

I don't think so. Expect that in a market with high vendor lock-in but that's not the case here. The market is extremely competitive and switching cost are near zero. Anthropic can't afford to pull shit like this and sacrifice quality.
ngruhn
·14일 전·discuss
[dead]
ngruhn
·18일 전·discuss
Is that news? Sounds like status quo.
ngruhn
·19일 전·discuss
> Sorry, but how was that murder successful?

Successful in winning over the public.
ngruhn
·19일 전·discuss
[flagged]
ngruhn
·22일 전·discuss
Also a lot of questions had "right answer" / "opposite of right answer" pairs. Just by identifying those you get to 50% probability.
ngruhn
·24일 전·discuss
I don't read the parent comment to take issue with the use case per se but the billion scans per month figure.

Surely, whatever this is giving you, getting a scan once a month must be plenty. They need a billion people to get a scan every month.
ngruhn
·27일 전·discuss
> I want to do offensive programming (I coined that term). You take risks, but you fix things quickly and ship.

Nice, I like the term too. But the paradigm is absolutely status quo in the industry. The thing is: with Gen AI the cost of "defensive programming" has gone way down, while the cost of (human) verification has gone way up. On the other hand, formal methods make verification cheap but come with massive implementation overhead (writing specs, types, proofs, and generally bending the implementation into a rigid framework). But Gen AI can automate all that laborious work. It's a match made in heaven.
ngruhn
·27일 전·discuss
You can't get rid of undefined. You can't ban null either. They're both used in the core language all over the place:

undefined camp:

- out of bounds array index access [1,2,3][42]

- non existent object key access {}.foo

- array.find(...) no result

- ...

null camp:

- document.querySelector(...) no result

- regex.match(...) no result

- ...
ngruhn
·28일 전·discuss
> a big attack surface

Wdym? At least web apps are sandboxed by default in contrast to native.
ngruhn
·28일 전·discuss
I agree. I'm not a game dev. I had a game idea and vibe-coded it with Claude. I kinda got what I had in mind but the game is just not engaging. I don't even know what to prompt. I tried "how could the be made more engaging" but no good ideas are falling out. I just have a lack of intuition for this kind of development. And Claude doesn't help.
ngruhn
·30일 전·discuss
might as well be in Skyrim
ngruhn
·30일 전·discuss
Kinda crazy if you think about it. In my home town of Cologne, if you dig anywhere you either find roman ruins or WW2 bombs. ~30 bombs last year. I've been evacuated so many times.
ngruhn
·지난달·discuss
Youtubers are using something as sources. Someone still has to do the journalism. Not sure YouTubers can replace that.
ngruhn
·지난달·discuss
maybe mark it as spam? If you don't have an account anyway, you don't miss anything. Also maybe they get punished a bit by being ranked as spam
ngruhn
·지난달·discuss
Despite having a NYT audio subscription, I continue to listen to the ads in a third party podcast app just so I don't have to use their god damn app.
ngruhn
·지난달·discuss
> I'm intensely skeptical about anything Anthropic says, because they are so incented to make their products seem dangerous

OpenAI, Google, etc. are not using "that strategy". I do believe that people at Anthropic genuinely care about AI safety. That's the main reason the company was founded. But I can imagine that idealism is eroding with new people and money flowing in.