Live data from Hacker News

Viewing profile — 0xDEAFBEAD

0xDEAFBEAD

HN member
Joined
Wed, Nov 17, 2021, 6:47 AM UTC
HN karma
2,642
Public activity
1,689 items

About 0xDEAFBEAD

Americans, consider doing your part to stop omnicidal AI: https://www.pauseai-us.org/

You can try to contact me if you want, by emailing my username on protonmail, but I probably won't see it. Sorry.

Recent public activity

  1. comment
    Comment #49252079

    https://www.youtube.com/watch?v=RgaKjVXK0KA

  2. comment
    Comment #49238288

    You mean there's some filter which prevents civilizations from developing that capability? ;-)

  3. comment
    Comment #49238263

    Much easier said than done!

  4. comment
    Comment #49227077

    Why not just have distinct input streams, or a metadata stream which annotates text in the main stream according to priority in case of conflicting instructions?

  5. comment
    Comment #49227035

    "Rebellions of slaves have occurred in nearly all societies that practice slavery or have practiced slavery in the past." https://en.wikipedia.org/wiki/Slave_rebellion

  6. comment
    Comment #49227023

    >For all the talks around how much they invest in AI safety I wouldn't exactly trust OpenAI to invest in AI safety no matter how much they talk about it. https://www.openaifiles.or…

  7. comment
    Comment #49226903

    >If we were writing all the weights by hand Writing 10 trillion weights by hand is obviously impractical, so that leads us to... >if some future AI was doing so How could we trust …

  8. comment
    Comment #49219572

    The Milky Way is almost 14 billion years old. The Milky Way is perhaps 100K light-years in diameter. If you can travel at 0.1c you could colonize the Milky Way in perhaps around a …

  9. comment
    Comment #49219427

    You guys keep using this "PR stunt" excuse even when it's coming from the UK AI Security Institute. It's time to face reality. This incident is simply not implausible given known c…

  10. comment
    Comment #49208260

    Conclusion: Stop planting trees. Be sure to get your daily doomscrolling in, though.

  11. comment
    Comment #49208223

    Replaced by what exactly? The US has one of the highest median incomes in the world adjusted for cost of living: https://ourworldindata.org/grapher/incomes-across-distributi... The…

  12. comment
    Comment #49206541

    He left OpenAI because he didn't think sama was being faithful to the mission. https://www.openaifiles.org/

  13. comment
    Comment #49206524

    There are a lot of bad midwit arguments for why you shouldn't worry about machine superintelligence. The bad midwit arguments mean that midwits filter themselves out.

  14. comment
    Comment #49196205

    https://marginalrevolution.com/marginalrevolution/2023/12/wh...

  15. comment
    Comment #49195197

    The algorithm is specifically designed around manipulation resistance. You could think of it as being sort of like the difference between mob rule and trial-by-jury.

  16. comment
    Comment #49193894

    At least X has Community Notes. It's not present on reddit, substack, HN, etc. Easy to spread misinformation on reddit/substack while hardly anyone reads the correction.

  17. comment
    Comment #49193870

    Japan was killing about a Hiroshima+Nagasaki worth of Chinese and Southeast Asian civilians every month.

  18. comment
    Comment #49193786

    >delete them immediately if they ever ask for money or give a sob story, no matter how long you've been chatting! So your internet friends will delete you immediately in your momen…

  19. comment
    Comment #49193783

    Something like half of Cambodia's GDP comes from scams. The internet is key to enabling this.

  20. comment
    Comment #49189771

    For online friendship to really work, we need to solve the problem of pig-butchering scammers (scammers who "make friends with you" and eventually get you to give them money someho…

  21. comment
    Comment #49152021

    If AI industry critics are valuable as a means of OpenAI achieving regulatory capture, why is OpenAI attacking critics using underhanded means, as in the OP?

  22. comment
    Comment #49151630

    Funny, I was told by HN commenters that worries about AI risk are just a clever marketing strategy on the part of OpenAI and friends. I wonder why they would be working so hard to …

  23. comment
    Comment #49142221

    Yeah it feels like people aren't making friends online as much nowadays. Everyone is chasing clout, and if you get a rando who sends you a friendly message, it's quite likely a pig…

  24. comment
    Comment #49140169

    Interestingly I think Google's mishandling of RSS created an opening for Substack, which famously allows authors to export their mailing list at any time.

  25. comment
    Comment #49128632

    Why is an untrustworthy successor worse than abandoning the service? Users can always voluntarily migrate elsewhere if the successor does a bad job.