Live data from Hacker News

Viewing profile — tvirosi

tvirosi

HN member
Joined
Sat, Jul 03, 2021, 11:38 AM UTC
HN karma
117
Public activity
51 items

About tvirosi

No profile information was provided.

Recent public activity

  1. comment
    Comment #28246935

    I agree, they're awful. But on the other hand the horror market exists so somehow people enjoy seeing things like this (in certain moods).

  2. story
  3. comment
    Comment #28233095

    I admit that I never really understood this speech. Could you explain why it was significant to you and what it means?

  4. comment
    Comment #28232562

    Agree. The gifs are so distracting, I could barely get through it.

  5. comment
    Comment #28224983

    Or the 47k no fly number is just a lie

  6. comment
    Comment #28224307

    Where am I supposed to look in that pdf to understand that it isn't correct or credible? It is certainly true that the model has differentiable and thus optimizable outputs.

  7. comment
    Comment #28224005

    It might have to do with the output possibly being a probability vector as opposed to a binary hash. The whole thing is thus differentiable and optimizable (if a dog image was inco…

  8. comment
    Comment #28221303

    Probably the model weight files

  9. comment
    Comment #28221245

    I've tried inference on the python version of onnx and it usually varies between hitting a OOM limit (while with TF it works fine) to being an order of magnitude slower. Even if th…

  10. comment
  11. comment
    Comment #28219891

    Makes me wonder if there's a possibility of e.g. faces on fbi's most wanted being snuck into the dataset somewhere in the chain.

  12. comment
    Comment #28219867

    Onnx has really poor support by microsoft. I suspect they've basically abandoned it internally (their onnxjs variant is orders of magnitude slower than tfjs[1]). It's a good 'neutr…

  13. comment
    Comment #28219223

    Or for criminals to generate perceptually similar illegal images that are no longer triggered as a 'bad' hash.

  14. comment
    Comment #28219201

    Really disgusting idea: I wonder if it's possible for someone to use this as a 'discriminator' in a GAN to configure a generator to recreate the CP this is trying to avoid distribu…

  15. comment
    Comment #28192192

    This might totally work and it's kind of impressive if it does. I'm still biased towards ultra skepticism towards all of this since the trustworthiness of all demos like this is co…

  16. comment
    Comment #28181999

    People are "allowed" to ignore drugs and nicotine but it's pretty hard and culture and laws help us constrain the lower bound will power you have to have in order to live a life. I…

  17. comment
    Comment #28181334

    I don't believe in this perspective. I think most people are sick of it (and would live healthier lives if they were allowed to focus more attention on their immediate surroundings…

  18. comment
    Comment #28140424

    I feel like they're just gonna keep pushing for this relentlessly and eventually the public will be too tired to fight back and then they'll win.

  19. comment
    Comment #28132590

    "code is law" unless those with concentrated power doesn't like the code

  20. comment
    Comment #28101484

    His disagreeability is probably correlated with his track record of innovation and if we cancel every outlier like that our ability to reach new plateaus will lag decades behind th…

  21. comment
    Comment #28101233

    Cool article! Those clips are from sweden right? :) Just curious if I spotted it right

  22. comment
    Comment #28098138

    Weird that the heads of apple seem to have redefined 'in true apple fashion' from just integrating well across the hardware and software layer to being about every part of the orga…

  23. comment
    Comment #28091268

    No need for a tone like that

  24. comment
    Comment #28087886

    Huh? Half precision floats speeds up machine learning training and inference enormously. It's weird to me to argue that we should sacrifice hours of sliced off waiting time in favo…

  25. comment
    Comment #28085192

    Not long until people will use deep learning style techniques to modify images to falsely trigger this alarm on harmless pictures and strew them randomly on the web.