Live data from Hacker News

Viewing profile — thatsadude

thatsadude

HN member
Joined
Wed, Mar 02, 2016, 4:54 PM UTC
HN karma
36
Public activity
79 items

About thatsadude

No profile information was provided.

Recent public activity

  1. comment
    Comment #49186963

    This is “google brain”

  2. comment
    Comment #48655151

    Funny to think about this when he got fired releasing a well-received product. On the other hand, Google released a subpar antigravity 2.0, full of bugs, almost unusable. The tech …

  3. comment
    Comment #48442795

    So are my DSP/Audio ML career, I spent decades acquiring the expertise. Now at 40, I don't know what to do if I lost my current job. Very sad, but from economic POV, I also don't n…

  4. comment
    Comment #48332692

    Nawh, they trained on test since Llama 2, no wonder.

  5. comment
    Comment #47233709

    On semantic VAD, I recommend https://github.com/pipecat-ai/smart-turn

  6. comment
    Comment #46872523

    I hope SpaceX succeeds

  7. comment
    Comment #46410506

    Nice read!

  8. comment
    Comment #45761352

    Has been a life changer for me.

  9. comment
    Comment #45208340

    See: https://arxiv.org/pdf/2411.03923

  10. comment
    Comment #45207719

    Same. I went to this thread and saw similar talking points brought up. Dangerous time for the US ahead.

  11. comment
    Comment #45203806

    [flagged]

  12. comment
    Comment #45203352

    [flagged]

  13. comment
    Comment #45192895

    Bad investment IMHO. Mistral was started by people who cheated on benchmarks with their Llama 1. It showed as they had the head start but fell far behind Gemini, DeepSeek and Qwen …

  14. comment
    Comment #42349936

    You can optimize prompt with MIPROv2 without examples (set the max number of examples to 0)

  15. comment
    Comment #42349928

    My go to framework. I wish we can use global metrics in DSPy, for examples, F1 score over the whole evaluation set (instead of a single query at the moment). The recent async suppo…

  16. comment
    Comment #41786729

    I’m disappointed. Nobel price for neural network architects that failed the test of time..

  17. comment
    Comment #41786643

    Well more deserving than the author of Restricted Boltzmann machine.

  18. comment
    Comment #41785363

    True, that's a yellow flag for me.

  19. comment
    Comment #41217068

    I had a few problems with DSPy: * Multi-hop reasoning rarely works with real data in my case. * Impossible to define advanced metrics over the whole dataset. * No async support

  20. comment
    Comment #40901421

    That plastic dish (parabolic reflector) is not cheap.

  21. comment
    Comment #40873685

    Recent ASR models are already robust to noise due to Spec augment and large-scale data. If you use these noise reduction services to remove noise, ASR models will have harder time …

  22. comment
    Comment #40412578

    20 years from now, the future generation will laugh at how delusional some tech guys think that "text generation could be and end to humanity".

  23. comment
    Comment #39356661

    You can buy a USB microphone array and create one yourself. - https://pysdr.org/content/doa.html - https://www.minidsp.com/products/usb-audio-interface/uma-16-...

  24. comment
    Comment #38347879

    Based on credits in Gpt3 and 4 papers, I think the team that follows Sam and Greg are the main drivers of the tech. Ilya is an advisor more or less.

  25. comment
    Comment #38346693

    Go read the gpt3 and gpt3 tech report and see for yourself.