adam_rida

13 ms·

1. 1.▲
   Google bid $10M for Spirit’s data. Our sensitivity analysis reaches $50M - $200M(tracerml.ai)

   1 points by adam_rida 21d ago|0 comments
2. 2.▲
   by adam_rida 1mo ago

   i typically run many coding agents side by side and i noticed i never read the wall of tool calls and text and tasks they do, i just want to know if 1) it’s still going 2) what is it currently doing and 3) what was its last message if it st
3. 3.▲
   Show HN: HUD, an open-source minimal terminal UI for ClaudeCode, Codex, OpenCode(github.com)

   26 points by adam_rida 1mo ago|1 comments
4. 4.▲
   by adam_rida 2mo ago

   i think that’s exactly the core of the question. There is meaningful decorrelation across models in some domains, like language for instance. but i agree that it’s much less obvious for reasoning. I don’t think more models is necessarily be
5. 5.▲
   by adam_rida 2mo ago

   thanks! adjusting it now
6. 6.▲
   by adam_rida 2mo ago

   good catch, on it too now!
7. 7.▲
   by adam_rida 2mo ago

   exactly, if you ensemble heavily small uncorrelated models (while each being expert on its task) you can get really interesting resutls. on agentic and coding what's make the problem even deeper is the granularity. how and when to use
8. 8.▲
   by adam_rida 2mo ago

   removing it now
9. 9.▲
   by adam_rida 2mo ago

   there are and this what you optimize for. ensemble learning has a long literature on this. you want models that have the most diverse pool of capabilities so they complement each other. in verifiable tasks or classification this is straight
10. 10.▲
    by adam_rida 2mo ago

    thanks to everyone for taking the time to try Echo and share feedback, this is precisely why i wanted to launch early. i am going to try to address a couple of topics that came up often: - i'll keep publishing stronger evals, includin
11. 11.▲
    by adam_rida 2mo ago

    looking at it now
12. 12.▲
    by adam_rida 2mo ago

    looking at it now
13. 13.▲
    by adam_rida 2mo ago

    The evaluator is public here: https://echo.tracerml.ai/eval/ It currently exposes 907 stored rows across seven benchmark families, with prompts, outputs, grades, and cost records. More benchmarks are coming soon. Echo
14. 14.▲
    by adam_rida 2mo ago

    Thanks for checking the individual rows. HumanEval+ is one small code slice, not the whole basis for the launch claim. The public evaluator currently contains 907 rows across seven benchmark families, and matched SWE-bench Verified and BigC
15. 15.▲
    by adam_rida 2mo ago

    You are right that the privacy wording was too broad. We are fixing it now so it states explicitly that Echo does not use customer prompts, files, chats, or outputs to train or fine-tune models. We are updating the matching Terms language a
16. 16.▲
    Show HN: Echo – Fable-level results at 1/3 the cost using open-weight models

    484 points by adam_rida 2mo ago|228 comments
17. 17.▲
    by adam_rida 3mo ago

    very cool
18. 18.▲
    by adam_rida 3mo ago

    thanks, apologies for the broken link This one should work: https://www.theregister.com/ai-and-ml/2026/06/14/ai-is-code-...
19. 19.▲
    by adam_rida 3mo ago

    Apologies, the submitted link seems broken. Actual link: https://www.theregister.com/ai-and-ml/2026/06/14/ai-is-code-...
20. 20.▲
    FTX's former Anthropic stake would be worth about $75B at today's valuation

    42 points by adam_rida 3mo ago|22 comments
21. 21.▲
    AI is code and can't be prompted into being smarter(theregister.com)

    15 points by adam_rida 3mo ago|16 comments
22. 22.▲
    Google sues alleged Chinese cybercrime operation over AI-generated scam texts(techcrunch.com)

    3 points by adam_rida 3mo ago|0 comments
23. 23.▲
    Nvidia raises RTX Pro 6000 Blackwell GPU pricing to $13,250(tomshardware.com)

    9 points by adam_rida 3mo ago|0 comments
24. 24.▲
    by adam_rida 3mo ago

    very interesting, arabic is a good reminder that text rendering is mostly solved for the scripts that shaped the defaults. The hard part is that typography, shaping, bidi behavior, font fallback, search, and the editor model all leak into e
25. 25.▲
    by adam_rida 3mo ago

    very cool
26. 26.▲
    by adam_rida 3mo ago

    Thanks for sharing! Happy to answer any questions
27. 27.▲
    by adam_rida 3mo ago

    This feels like the natural next phase. A lot of companies went from “use as much AI as possible” to suddenly realizing that not every call needs frontier-model intelligence. In production, many calls are just repeated classification, taggi
28. 28.▲
    by adam_rida 5mo ago

    Very cool!
29. 29.▲
    by adam_rida 5mo ago

    very interesting
30. 30.▲
    by adam_rida 5mo ago

    very cool!