nojs

16 ms·

1. 241.▲
   by nojs 1y ago

   This does not contradict the article - it may be true, and yet not significantly more productive, because of the increased review burden.
2. 242.▲
   by nojs 1y ago

   I’m surprised this article didn’t mention the LPC code [1] 1. https://youtu.be/jPhXVrp0_oI
3. 243.▲
   by nojs 1y ago

   Haha, he’s renamed it from “shit” to “repository” which makes the comments less funny
4. 244.▲
   by nojs 1y ago

   For people using these, is there an advantage to having the agent create PRs and reviewing these versus just iterating with Cursor/Claude Code locally before committing? It seems like additional bureaucracy and process when you could f
5. 245.▲
   by nojs 1y ago

   > I think the only productive use of time during those gaps is meditation Check out several copies of the repo and work on different branches concurrently.
6. 246.▲
   by nojs 1y ago

   > The only real thing I'm missing is the diff view across files You can commit checkpoints prior to each major prompt and use any IDE’s builtin visual diff versus last commit. Then just rebase when the task is done
7. 247.▲
   by nojs 1y ago

   It’s more of a ticking time bomb because it relies so heavily on upstream model providers who all have competing products, particularly Claude Code.
8. 248.▲
   by nojs 1y ago

   > over half of the eggs broke when dropped vertically from an 8-millimeter (31-inch) height Something went wrong with the units I think
9. 249.▲
   by nojs 1y ago

   > it's that the thought that went into making the prompt is, inevitably, more interesting/original/human than the output the LLM generates afterwards I think you are overestimating the people who submit this slop. It’s mor
10. 250.▲
    by nojs 1y ago

    That’s what they’re evidently trying to do, but it means they’re moving too slow and the AI integration feels like a bolted on afterthought (hence the reaction to failed features like this). I am a lifetime user of PyCharm but the reality i
11. 251.▲
    by nojs 1y ago

    They’re in a difficult position because half their users want more AI but the other half complain loudly when it’s forced on them. Cursor is beating them because they can deeply embed AI everywhere without worrying about this.
12. 252.▲
    by nojs 1y ago

    This works well but only if you eyeball the tests and edit them a bit in my experience. Otherwise it gets lazy and makes them trivial to pass. Also, you’ve often gotta explicitly tell it not to hardcode test cases in the solution to make th
13. 253.▲
    by nojs 1y ago

    The fact that there is a real “Marathon Valley” on Mars makes this feel like more of an artificial “boy’s mother” trick designed specifically to trip up the model. It seems reasonable to assume that’s what you meant.
14. 254.▲
    by nojs 1y ago

    I think the author is missing the point of why these forecasts put so much weight on software engineering skills. It’s not because it’s a good measure of AGI in itself, it’s because it directly impacts the pace of further AI research, which
15. 255.▲
    by nojs 1y ago

    This is so good. Reminds me of The Office. I love how bad the other examples are.
16. 256.▲
    by nojs 1y ago

    Claude
17. 257.▲
    by nojs 1y ago

    Incidentally what is the best way to cursify jetbrains at the moment?
18. 258.▲
    by nojs 2y ago

    https://xkcd.com/927/
19. 259.▲
    by nojs 2y ago

    Funny that it literally begins with “Follow @index@simone.org via your Fediverse account (like Mastodon, Pixelfed, Flipboard, Wordpress, Writefreely, Threads, or BlueSky.)” Or is that just helpfully letting us know of something related that
20. 260.▲
    by nojs 2y ago

    > People’s names are all mapped in Unicode code points Curious how the author recommends we avoid this one
21. 261.▲
    by nojs 2y ago

    > Hallucinated methods are such a tiny roadblock that when people complain about them I assume they’ve spent minimal time learning how to effectively use these systems—they dropped them at the first hurdle. If you’re writing code in Pyth
22. 262.▲
    by nojs 2y ago

    *the fountainhead
23. 263.▲
    by nojs 2y ago

    I used to do it every morning in a cold city (water was around 10°C) and it made me feel less cold throughout the day, not more. That was one of the more obvious effects.
24. 264.▲
    by nojs 2y ago

    > I hate it when HN people completely dismiss anecdotes and want research papers for everything. You can point them to this: https://pmc.ncbi.nlm.nih.gov/articles/PMC300808/
25. 265.▲
    by nojs 2y ago

    > This piece is nothing but an advertisement for Dan Gilbert. For an outlet that “thanks to our reader-funded model, what we cover isn’t dictated by the algorithms of the tech titans” they did manage to cram a surprising number of ads in
26. 266.▲
    Effects of two firefighting chemical formulations (2022)(sciencedirect.com)

    2 points by nojs 2y ago|0 comments
27. 267.▲
    by nojs 2y ago

    This may be of interest: https://amp.abc.net.au/article/103610270
28. 268.▲
    by nojs 2y ago

    Time for http-over-whatsapp
29. 269.▲
    by nojs 2y ago

    Cool project. Note that you can force structured output now instead of asking for json: https://platform.openai.com/docs/guides/structured-outputs
30. 270.▲
    by nojs 2y ago

    Alternatively if he predicted the probability of eruption every day as 0.01%, he’d be a perfect forecaster!