yawnxyz

17 ms·

1. 331.▲
   by yawnxyz 2y ago

   I’ve always thought of them as “official datasets” - is that right? With GitHub anyone can just chuck something up there with not much forethought, but not so much with DataCite right?
2. 332.▲
   by yawnxyz 2y ago

   I’ve been a bioinformatics engineer for a while. Publishing is narrative building, and wades through lots and lots of experimental data. Most of those aren’t super interesting. It’s like let’s say you record everything that’s ever been said
3. 333.▲
   by yawnxyz 2y ago

   I wonder if this only works for stories with narratives or will it break down for other kinds of reports? I guess even research papers have "narratives"
4. 334.▲
   by yawnxyz 2y ago

   reminds me of this Aussie cleaning company's website that forced you to create an account to take an order. With a couple of clicks on the web app, you'd encounter a bug... and then you can see every single person's orders, e
5. 335.▲
   by yawnxyz 2y ago

   Wow incredible, could you give a small glimpse in how you collected/normalized/vectorized all the different reviews + information? Just lots of scraping + chucking it into Pinecone etc? > Me: Give a list of the new and old gene
6. 336.▲
   by yawnxyz 2y ago

   don't forget the time everyone's Google Drive photos got corrupted
7. 337.▲
   by yawnxyz 2y ago

   At least with Cloudflare they'll get back to you in five minutes as soon as you complain on HN
8. 338.▲
   by yawnxyz 2y ago

   recently switched all our mail to FastMail and that took 1ish hour... I was ready to spend a couple of days to wrestle switching mail providers. Guess they really live up to their promise
9. 339.▲
   by yawnxyz 2y ago

   at least Rick Rubin admits he doesn't play any instruments
10. 340.▲
    by yawnxyz 2y ago

    > Niching down, if you work in operations at a <50 person startup or SMB and your company relies on a Postgres or MySQL database, Sourcetable is an affordable reporting tool with turnkey data infrastructure that doesn’t require code o
11. 341.▲
    by yawnxyz 2y ago

    hmmmm I consider myself an introvert and I liked it. Maybe I'm not as introverted as I thought...
12. 342.▲
    by yawnxyz 2y ago

    is this better than the Youtube-generated transcript / captions you'd get from something like https://github.com/Kakulukian/youtube-transcript ?
13. 343.▲
    by yawnxyz 2y ago

    try a community sport like basketball or kickball and find a team that likes to get food/drinks after. So fun
14. 344.▲
    by yawnxyz 2y ago

    I've been stuffing JS and API calls into Google Sheets cells and for small use cases its great! When you get very complicated flows, of one cell pointed to another cell, it becomes a massive mess. It's hard to debug and there&#x27
15. 345.▲
    by yawnxyz 2y ago

    I think knowing what part of the knowledge base to delete — to get to adequately small reasoning model — is the hard part. Doesn't "reasoning" rise from the knowledge? How much of a brain can you cut away before you affect th
16. 346.▲
    by yawnxyz 2y ago

    in the age of AI and Cursor, I make my function name as expressive as I can, and I make sure to add a couple of lines of comments, either generated or manual. It makes it way easier to send these into Claude (it seems, at least). I hope the
17. 347.▲
    by yawnxyz 2y ago

    haha I haven't tested it if it's efficient or anything, I just put it together as a pipeline on val.town and I've been using it for parsing. Shouldn't take more than 5 minutes to put together w/ Claude tbh
18. 348.▲
    by yawnxyz 2y ago

    I think the author did a great job communicating that some parts of software are fully open, and a few other pieces of code / repost are not. I like this way better than software with complicated licensing schemes, where you can only u
19. 349.▲
    by yawnxyz 2y ago

    I found that reducing html down to markdown using turndown or https://github.com/romansky/dom-to-semantic-markdown works well; if you want the AI to be able to select stuff, give it cheerio or jQuery access to navigate
20. 350.▲
    by yawnxyz 2y ago

    yeah usually much better results getting a list of ~10 fuzzy keyword search results, 10 semantic/embeddings results, and using something like Cohere rerank (or just a cheap GPT model) to choose the best 5-10 results from the pile
21. 351.▲
    by yawnxyz 2y ago

    Isn't that what happened to Evergrande
22. 352.▲
    by yawnxyz 2y ago

    Is it possible to build something like Canva / Figma in Repaint?
23. 353.▲
    by yawnxyz 2y ago

    wait, so if WordPress is migrating 500M blogs to Wordpress[1], does this mean essentially we'll have easy access to all tumblr blogs' history? [1] https://arstechnica.com/gadgets/2024/08/tumblr-migra
24. 354.▲
    by yawnxyz 2y ago

    It's so clever to just pull from Wayback Machine rather than scrape the site itself. Never even thought of that
25. 355.▲
    by yawnxyz 2y ago

    Every time I ask for cafe recommendations, half of its suggestions are listed as "permanently closed" on Google Maps :/
26. 356.▲
    by yawnxyz 2y ago

    from having built a bunch of tiny indie games in a prior life (in Flash :P), time = state, and time = world and depending on what games we're building, either we record actions as state changes (eg chess / turn based games: if no
27. 357.▲
    by yawnxyz 2y ago

    As a non-LLM, I got so confused with Svelte 5's runes, that I just avoided it altogether by not touching it haha
28. 358.▲
    by yawnxyz 2y ago

    Claude gave me something similar, except these were both used, and somehow global variables, and it got confused about when to use which one. Asking it to refactor / fix it made it worse bc it'd get confused, and merge them into a
29. 359.▲
    by yawnxyz 2y ago

    https://archive.is/oAFxn
30. 360.▲
    The Canadian dream is on life support(theglobeandmail.com)

    13 points by yawnxyz 2y ago|10 comments