HackerLangs
トップ新着トレンドコメント過去質問紹介求人

high_byte

863 カルマ登録 6 年前

投稿

I don't review my code anymore

twitter.com
3 ポイント·投稿者 high_byte·28 日前·3 コメント

Ask HN: How to deal with App Store rejections?

4 ポイント·投稿者 high_byte·28 日前·3 コメント

Ask HN: Xcode with Tailscale for Remote Coding

2 ポイント·投稿者 high_byte·先月·0 コメント

[untitled]

1 ポイント·投稿者 high_byte·2 か月前·0 コメント

What are some vibecoding horror stories?

6 ポイント·投稿者 high_byte·3 か月前·2 コメント

How Much Does AI Cost?

ai-price.netlify.app
2 ポイント·投稿者 high_byte·8 か月前·1 コメント

Ideas for Better Programming Tools

emilprogviz.com
3 ポイント·投稿者 high_byte·9 か月前·1 コメント

コメント

high_byte
·3 日前·議論
this is definitely a backdoor, not necessarily that they use it to infiltrate users but definitely they put them at risk.

reminds me of a bug I found in some tplink router it compared passwords of 3 different users but that table was empty so basically 15 NULL bytes would log you in as admin lol
high_byte
·3 日前·議論
what if it contained a zero day for tesseract and the script you thought you got is just a throwaway
high_byte
·7 日前·議論
those spikes in march and june? war with iran. interesting...
high_byte
·8 日前·議論
my experience with ffmpeg scene detection is that it's flaky. it works, sometimes, but not reliable by any means
high_byte
·10 日前·議論
would be nice to be able to simply get the list instead of having an ominous heatmap
high_byte
·11 日前·議論
this was absolutely fascinating to read
high_byte
·16 日前·議論
that's sort of what skills are.

create a skill that transformed the json then "/<skill> <prompt>"
high_byte
·16 日前·議論
code is like assembly now.

in the olden days (pre-LLMs) we would write high-level code.

the entire layer was high-level code and rarely would we ever need to peak into the assembly:

writing, debugging, architecting, reviewing, testing - all were done in the high-level language layer.

---

welcome to present day:

since we don't write code - we write intents, we also shouldn't review code either - we should review intents.

I don't review my code anymore. I ask the agent to generate markdown docs, graphviz diagrams, changelogs, audit reports, etc. I only review that.

I also ask it to write test and evaluate by whether the tests passed or not. I don't need to peak into the tests code - I can also ask plain english, pseudocode, control flow graph, whatever it is I want.

I can ask it to find errors or missing tests and improve that too!

code is like assembly now.

rare are the cases you would need to peak into that level.
high_byte
·17 日前·議論
this is amazing exactly what I always wanted

how did this even get approved by app store? that's actually surprising
high_byte
·24 日前·議論
interesting attack vector

shame it's not a real device so the surface is limited to bash only

I wonder what software might be vulnerable to this attack surface
high_byte
·26 日前·議論
rick rubin says otherwise
high_byte
·26 日前·議論
of course
high_byte
·27 日前·議論
classic HN
high_byte
·28 日前·議論
I am the founder and ceo
high_byte
·28 日前·議論
"the code review: if you can’t review code, you can’t direct AI tools. You cannot lead a team. You cannot even trust your own work."

I would revise that a bit

since we don't write code, we write intents - we shouldn't review code either, we should review intents.

I don't review my code anymore. I ask the agent to generate markdown docs, graphviz diagrams, changelogs, audit reports, etc. I only review that.

I also ask it to write test and evaluate by whether the tests passed or not. I don't need to peak into the tests code - I can also ask plain english, pseudocode, control flow graph, whatever it is I want.

I can ask it to find errors or missing tests and improve that too!

so yea - code is like assembly now. rare are the cases you need to peak into that level.
high_byte
·28 日前·議論
even if the most powerful ai is open source and let's even assume runs on consumer hardware - in the end data is the real moat.

if it can access private data it will necessarily have more power.
high_byte
·28 日前·議論
honestly I don't think I've ever seen D codebases??
high_byte
·28 日前·議論
on the other hand steve jobs would've called Qt human-slop

guess it's a matter of taste
high_byte
·29 日前·議論
I am using cursor on auto and I got the exact same experience.

installed quartz, used accessibility and screen recording api, all that.

initially it managed to do it on another desktop space somehow, opening safari in the background without me even noticing. but then it actually started using my own mouse while I was using it lol
high_byte
·先月·議論
all this for 2%