Live data from Hacker News

Viewing profile — danielcampos93

danielcampos93

HN member
Joined
Thu, Jun 06, 2013, 11:39 PM UTC
HN karma
449
Public activity
116 items

About danielcampos93

Perfection is the enemy of good https://spacemanidol.github.io/

Recent public activity

  1. story
  2. comment
    Comment #45966336

    mums the word on Flash?

  3. comment
    Comment #45966030

    I would love to know what the increased token count is across these models for the benchmarks. I find the models continue to get better but as they do their token usage also does. …

  4. comment
    Comment #45386555

    It's pretty wild that Brave's terms of service state as much, considering their search API is entirely derived from storing the results of other search systems. https://support.bra…

  5. story
  6. comment
    Comment #45171624

    I think you would be very surprised by the number of customers who don't care if the embeddings are SOTA. For every Joe who wants to talk GraphRAG + MTEB + CMTEB and adaptive rag t…

  7. story
  8. story
  9. story
  10. comment
    Comment #43323711

    they also 10xed their revenue. 24 seems low for someone that pulled that off.

  11. comment
    Comment #43161193

    GPT-4o as a judge to evaluate the quality of something which gpt4o is not inherently that good at. Red flag.

  12. comment
    Comment #43082401

    does it do the full 100? In my experience anything around many items that needs to be exhaustive (all states, all fortune 100) tends to miss a few.

  13. comment
    Comment #43082379

    It's not free because it's cheap for them to run. It's free because they are burning that late-stage VC dollars. Despite what you might believe if you only follow them on twitter t…

  14. comment
    Comment #42215634

    This works better because it gives a secondary set of conditions for which the decoder (generating text) is conditioning its generation. Assume instead of their demo you are doing …

  15. comment
    Comment #42175524

    https://www.youtube.com/watch?v=ecnS1Ygf0o0 I've been waiting for the chance to use this

  16. comment
    Comment #41649645

    Not mentioned in their blog posts but on the model cards on huggingface: "Molmo 72B is based on Qwen2-72B and uses OpenAI CLIP as vision backbone. Molmo-72B achieves the highest ac…

  17. comment
    Comment #41626995

    That is a funny was of explaining that they scrape google.

  18. story
  19. comment
    Comment #40626513

    It needs updating to include you.com, perplexity, etc. Most of those are google reskins/emulators but they are there non the less

  20. comment
    Comment #40610809

    This seems to be a feature that execs wanted, and people find creepy, and no one has the gumption to push back on the exec request.

  21. comment
    Comment #40457946

    More countries have nukes than companies have actual indexes

  22. comment
    Comment #39973539

    If you only emulate the teacher can you ever become an master?

  23. story
  24. comment
    Comment #39673723

    I feel like I have been hearing that since V1 TPU. I think Google is the perfect solution because they are teams whose job is to take a model and TPUify it. Elsewhere there is no t…

  25. comment
    Comment #39594438

    Claude's answers sometimes fill the niche of 'management consultant'