djoldman

22 ms·

1. 391.▲
   Chicago-based Amtrak trains are running out of seats(hsrail.org)

   1 points by djoldman 1y ago|0 comments
2. 392.▲
   by djoldman 1y ago

   Faster? At making what? Pipe dev/null and you'll get a lot of stuff fast. What if someone came out with a study saying they had a tool to make the fastest doctors and lawyers? You'd say that doesn't even make sense, what
3. 393.▲
   by djoldman 1y ago

   Agreed. Data curation is barely on the radar of most non-tech industries. Even in tech, it's rare to have any meta data. This is a huge blocker to many efforts. Someone somewhere has to go through every table and field and document whe
4. 394.▲
   by djoldman 1y ago

   > Next, you cover the whiteboard in boxes, arrows, and at least one redundant Kubernetes cluster. Add a message queue, Kafka obviously, regardless of whether you need one. Sprinkle in some microservices because monoliths are for peasants
5. 395.▲
   by djoldman 1y ago

   > Escalate risk of physical harm to others for human review. When we detect users who are planning to harm others, we route their conversations to specialized pipelines where they are reviewed by a small team trained on our usage policie
6. 396.▲
   by djoldman 1y ago

   This is quite the piece. One section that stands out is: > It’s no coincidence that as grades have gone up, civic friendship between people with different views and backgrounds has gone down, both on campuses and in the nation at large.
7. 397.▲
   by djoldman 1y ago

   This is a non-story. Stocks are up ~9% over the last 6 months and are marginally off of all time highs.
8. 398.▲
   by djoldman 1y ago

   https://archive.ph/joSRX
9. 399.▲
   by djoldman 1y ago

   Details on page 10 here: https://prod-i.a.dj.com/public/resources/documents/WSJNORCJu... Including previous values: 25% July 2025 29% July 2024 28% March 2023 27% May 2022 Interesting other measurement: Neith
10. 400.▲
    by djoldman 1y ago

    https://archive.ph/wMPIK
11. 401.▲
    by djoldman 1y ago

    Previously: https://news.ycombinator.com/item?id=9611829
12. 402.▲
    Rearchitecting GitHub Pages (2015)(github.blog)

    35 points by djoldman 1y ago|1 comments
13. 403.▲
    by djoldman 1y ago

    TIL about the SafeType keyboard. Wild. https://en.wikipedia.org/wiki/Ergonomic_keyboard#/media/File...
14. 404.▲
    by djoldman 1y ago

    Yep, this would be a big problem. We'd have to have alternate methods as well.
15. 405.▲
    by djoldman 1y ago

    Americans don't care. Journalism and news compete with entertainment. Everyone needs to eat. Lots of jobs have gone away and lots have been created. Just because industries shrink doesn't mean they should be bailed out or supporte
16. 406.▲
    by djoldman 1y ago

    Ok. If Americans don't value traditional journalism enough to pay what it may cost, it will go away.
17. 407.▲
    by djoldman 1y ago

    Americans already, and increasingly, report getting a good chunk of their news from social media: https://www.pewresearch.org/journalism/fact-sheet/social-med... Folks who want more traditional journalism will pay
18. 408.▲
    by djoldman 1y ago

    > ChatGPT, Google, and its competitors are rapidly diverting traffic from publishers. Publishers are fighting to survive through lawsuits, partnerships, paywalls, and micropayments. It’s pretty bleak, but unfortunately I think the situa
19. 409.▲
    by djoldman 1y ago

    KSVD Algorithm: https://legacy.sites.fas.harvard.edu/~cs278/papers/ksvd.pdf
20. 410.▲
    OEmbed(oembed.com)

    3 points by djoldman 1y ago|0 comments
21. 411.▲
    by djoldman 1y ago

    Yep. 10 times worse.
22. 412.▲
    by djoldman 1y ago

    In a world with no friction, one would want tariffs to apply to all goods. The de minimis exemption dealt with the fact that it was a PITA to do all the paperwork etc. that accompanies a tariff. Obviously the main problem is that tariffs do
23. 413.▲
    by djoldman 1y ago

    From table 3 it appears that Deepseek R1 has the highest eval scores. It's a 607B model vs 405B, so obviously "larger"
24. 414.▲
    by djoldman 1y ago

    Yes! Thanks. In particular one might use "[is] instance of": https://m.wikidata.org/wiki/Property:P31 So we would take in all items with that property to make the graph. Although we might have to deal with mu
25. 415.▲
    by djoldman 1y ago

    Sorry, the "web" isn't "open" and hasn't been for a while. Most interaction, publication, and dissemination takes place behind authentication: Most social media, newspapers, etc. throttle, block, or otherwise t
26. 416.▲
    by djoldman 1y ago

    Anyone know of work to automatically create or derive a taxonomy from wikipedia? This would be a directed acyclic graph like schema.org
27. 417.▲
    by djoldman 1y ago

    Here's the SynthID paper: https://www.nature.com/articles/s41586-024-08025-4.pdf https://www.nature.com/articles/s41586-024-08025-4
28. 418.▲
    by djoldman 1y ago

    This type of question, if the company truly wants an answer to it, should only be asked at the end of an interview process. Most people have no idea what a job is actually like before they have the job, much less during the interview proc
29. 419.▲
    by djoldman 1y ago

    I would appreciate it if you could describe the interview process in more detail for the position you accepted.
30. 420.▲
    by djoldman 1y ago

    just turn off JS.