Live data from Hacker News

Viewing profile — zmmmmm

zmmmmm

HN member
Joined
Mon, Feb 15, 2010, 2:06 AM UTC
HN karma
21,273
Public activity
4,019 items

About zmmmmm

No profile information was provided.

Recent public activity

  1. comment
    Comment #49256946

    > any better solution is either much more difficult or impossible. I just don't buy that. I think it is perfectly feasible to mandate device makers support parental controls that i…

  2. comment
    Comment #49253247

    the weird thing about it all is that age is a bad proxy in the first place for a child's level of safety in interacting with online content. Some kids are super mature at 11 and ot…

  3. comment
    Comment #49251267

    the big thing containers don't allow is for the agent to run and use docker itself without compromising the host I'm not sure where "vast majority" cuts in but I would say a huge n…

  4. comment
    Comment #49242129

    Meta knows how to win back developer's hearts .... let's see if they have the goods

  5. comment
    Comment #49240505

    It feels to me that Anthropic is pushing hard here to try and establish acceptance of a tolerance of risk such that their tools can run fully automated without supervision, but dir…

  6. comment
    Comment #49240137

    it's running a full VM so the agent can eg: run docker commands safely etc

  7. comment
    Comment #49238649

    It's funny / ironic how the whole agentic development boom has taken off in terminal environments, but now the orchestration / IDE layer seems to be all Electron / GUI apps. Any su…

  8. comment
    Comment #49237383

    > I definitely feel like there can be a pipeline where one extreme belief may lead to others Once you accept a premise that majority consensus of well informed people can be drasti…

  9. comment
    Comment #49227000

    Tobacco company says "we are launching a new product that will cause cancer and kill people. It's highly addictive so we expect widespread uptake. We think it is crucial that regul…

  10. comment
    Comment #49226970

    > If anything, I want these models to be less persistent at their focus of completing their goal I think it's honestly a slightly ugly form of benchmaxxing - they are desperate to …

  11. comment
    Comment #49226362

    I will say here that there are meaningful jobs that especially tech people can engage with. There are gigantic unsolved problems. Climate change, cancer, thousands of rare disorder…

  12. comment
    Comment #49226295

    On top of that, it seems the WFO crowd simply can't help exploiting it to their advantage once they are there. They get less work done, but they hoard higher level decisions about …

  13. comment
    Comment #49216934

    it's great but we need a multi-modal model of this quality and price to truly declare victory. But it makes me quite curious, how a text-only model can do so well on ARC-AGI-2 bein…

  14. comment
    Comment #49216767

    I started with the Lex Fridman podcast [0] and then I went from there exploring further with reading mostly online material interweaved with discussions with ChatGPT. It is actuall…

  15. comment
    Comment #49216704

    definitely But then it's also true the real arc of these things is only properly visible in retrospect. For example I'm sure historians looking back would include the Iraq war as o…

  16. comment
    Comment #49216662

    Some important context - it appears it applies to submissions from the community but possibly not to core developers: > Contributions in the OpenJDK Community must not include cont…

  17. comment
    Comment #49216603

    But there are no conclusions. It literally says: > Whether human contributions to AI-generated outputs are sufficient to constitute authorship must be analyzed on a case-by-case ba…

  18. comment
    Comment #49209569

    I read up recently on the history of the fall of the roman empire and it was fascinating how eerily similar it was to the current state of the US. Overwhelmed by corruption and war…

  19. comment
    Comment #49203201

    The fact that the Chinese models have caught up on benchmarks suggests to me that its likely we will start to transition now into much more of a brand war. It will be subjective qu…

  20. comment
    Comment #49192810

    Muse code is more interesting than the new model in that it seems to natively ship with an orchestrator / subagent pattern built in. Curious if it works well in practice compared t…

  21. comment
    Comment #49190956

    I don't think it's at all certain this won't land back on the same unit economics as the old way. The cost of serving a user doesn't have to be free - it never has been - it just h…

  22. comment
    Comment #49167760

    To me there are different motivations to include imagery and some of them are just fine with AI generation. If the imagery actually goes to the substance of helping to illustrate o…

  23. comment
    Comment #49165071

    Sadly while the FLOPS are increasing nicely, total graphics memory is stalled in consumer cards by comparison.

  24. comment
    Comment #49162132

    There's a growing and fascinating divide between people who see LLMs as more of a "bicycle for the mind" in the vein of Jobs vs those who see them as whollly supplanting the role o…

  25. comment
    Comment #49154011

    The outwardly measurable productivity increase will come but over a time scale of years, and most of that will be in barely measurable increments. Nobody wants to hear that but it'…